Sustainable Phosphate Recovery Using Novel Ca–Mg Bimetallic Modified Biogas Residue-Based Biochar

Abstract

Elevated phosphorus levels in aquatic ecosystems have been identified as a critical driver of eutrophication processes, necessitating resource-recovery remediation strategies. Adsorption techniques show particular promise for nutrient recovery due to their selective binding capacities and operational feasibility. In this study, the Mg- and Ca-modified biogas residue-based biochar (Ca-Mg/BC) was successfully prepared via a “bimetallic loading-pyrolysis” modification strategy. The optimum temperature for the calcination of the material and the salt solution impregnation concentrations were determined experimentally through optimization of the synthesis conditions. Structural and chemical analyses of Ca–Mg/BC demonstrated that the material contains MgO and CaO. The specific surface area of Ca–Mg/BC was 8.49 times higher than that of the unmodified biochar (BC). The optimized Ca–Mg/BC achieved 95% phosphate removal rate (157.13 mg/g adsorption capacity). FTIR and XPS characterization results indicated the importance of Ca/Mg loading in phosphate capture. MgO and CaO were mainly loaded on the surface of the material and adsorbed phosphate through a chemical reaction. Crucially, the phosphate-laden biochar exhibited potential as a nutrient-enriched soil amendment, opening the material loop from wastewater treatment to agricultural applications. This sustainable strategy simultaneously addresses water pollution control and sustainable development, providing environmentally benign solutions compatible with industrial effluent treatment and sustainable agriculture practices.

1. Introduction

The rapid expansion of biogas projects, while advancing renewable energy production, has precipitated challenges in sustainable waste management. Biogas slurry, a nutrient-rich byproduct of livestock manure, exhibits an annual production exceeding 400 million tons, with total phosphorus concentrations typically spanning from 30 to 200 mg/L. Owing to its high turbidity, intense coloration, and resistance to degradation, this effluent poses substantial environmental risks. As excessive phosphorus constitutes a primary driver of eutrophication, its discharge disrupts aquatic ecosystems but also potentially threatens human health through bioaccumulation [1]. Thus, the development of efficient and economically viable phosphorus-removal technologies for biogas slurry has become imperative to address agricultural non-point source pollution and advance the green transformation of agriculture under the “dual-carbon” strategy.

Biochar, a porous carbonaceous material produced by biomass at a high temperature (300–900 °C), offers a sustainable solution for phosphorus management [2,3]. Biochar has high porosity, a high specific surface area, and abundant functional groups, so it can be used as the adsorbent to adsorb phosphate in water [4]. However, conventionally, pyrolyzed biochar is limited by its functional group density, specific surface area, and anion exchange capacity. Notably, unmodified pristine biochar usually exhibits a poor phosphate-adsorption ability and may even release phosphate into the water [5,6]. Chemical modification represents an effective way to enhance the surface structure of biochar, enrich the type of functional groups, and improve the adsorption capacity for phosphate [7]. Studies demonstrate that cationic modifiers (such as Mg²⁺, Al³⁺, and Ca²⁺) substantially improve the phosphate adsorption [8,9,10]. The adsorption capacity of phosphate by calcium-modified biogas residue biochar is 154.18 mg/g, which is substantially higher than that of the original biochar [11]. Although these single-metal modified biochars have achieved the efficient adsorption of phosphorus, single-metal modified systems remain constrained by performance limitations in complex aqueous environments.

In contrast, bimetallic modification has garnered increasing attention due to its synergistic effects for superior nutrient recovery. Research has shown that using Mg/Fe bimetallic oxide-modified straw biochar can achieve a phosphate adsorption capacity of 206.2 mg/g [12], demonstrating the superiority of the bimetallic system. The enhanced performance stems from the richer functional group composition and stronger intermolecular interactions of bimetallic oxides compared to their monometallic counterparts [13]. Specifically, Ca-based materials can preferentially immobilize phosphate through their strong precipitation ability [14], whereas Mg-based materials offer dynamic active site replenishment due to their high mobility [15]. Combining these characteristics could transcend the limitations of single-metal systems. Moreover, the inherent Fe content in biogas residue provides a foundation for developing multimetal-loaded biochars. Nevertheless, achieving effective multimetal loading on biogas residue biochar remains a research challenge. To address this, based on this, this study proposes a novel “bimetallic loading-pyrolysis” modification strategy featuring CaO–MgO synergy through binary acetate impregnation and thermal modification. This modification process optimizes the pore structure while mitigating the pore collapse associated with conventional single-metal modification. Furthermore, the complementarity charge properties and co-precipitation effects of Ca/Mg ions enhance phosphate chemisorption, providing a new path for developing interference-resistant, high-performance adsorption-phosphorus fixation materials.

In this study, a sustainable approach was developed by preparing Mg–Ca bimetallic-modified biochar from biogas residue for phosphate recovery, creating a circular solution. The research included the following: (i) optimization of preparation conditions for waste-derived Mg–Ca bimetallic-modified biochar; (ii) evaluation of nutrient recovery performance under varying conditions (dosage, coexisting ions, pH) in both synthetic and actual biogas slurry, complemented by isotherm, kinetic and thermodynamic analyses; (iii) comprehensive characterization (SEM, SEM-EDX, FTIR, XRD, XPS) to elucidate the adsorption mechanisms enabling efficient phosphorus recycling; (iv) assessment of the fertilizer potential of phosphate-saturated biochar through sustainable agriculture pot experiments.

2. Materials and Methods

2.1. Preparation of Materials

All reagents were bought from Aladdin Industrial Company in Shanghai, China. After drying, the original biogas residue was ground to 100 mesh and named biogas residue powder (BR) (China/Shandong); 2 g of BR was added to 100 mL of the mixed solution. The mixed solution was composed of 0.3 mol/L Mg(CH₃COO)₂ and 0.3 mol/L Ca(CH₃COO)₂. The mixture of the biogas residue powder and salt solution was stirred at 150 rpm for 5 h. Following this procedure, the modified biogas residue mixture was dried in an oven at 65 °C for 12 h and subsequently ground to obtain the modified biogas residue composite material.

The modified biogas residue composite and the initial BR were placed in the porcelain boat of the tubular furnace. High-purity argon (99%) was introduced into the tubular furnace, and the temperature was raised to 700 °C at 5 °C/min. After calcining at a constant temperature for 2 h, those samples were taken out from the tubular furnace once the temperature had reduced to room temperature. The materials were ground and sealed, the modified biogas residue-based biochar was named Ca–Mg/BC, and the unmodified biogas residue-based biochar was named BC.

A phosphate solution with a concentration of 1000 mg/L was prepared by dissolving an appropriate amount of KH₂PO₄ in deionized water and diluting it to a volume in a volumetric flask. Subsequently, the prepared reserve solution was diluted to the concentration required for subsequent experiments.

2.2. Batch Adsorption Experiments

The effects of the initial pH, concentration, and the existence of other ions on the adsorption properties of the Ca–Mg/BC were determined via the isothermal adsorption method. In this experiment, a 100 mg/L phosphate adsorption solution was used, and Ca–Mg/BC (15 mg) was added to 30 mL of the adsorption solution and shaken well at 150 rpm, and the supernatant was filtered through the 0.45 μm filter.

The biochar dosage experiment was performed to determine the optimal dosage of the Ca–Mg/BC material in the experiment. A series of dosages of Ca–Mg/BC (0.17, 0.34, 0.5, 0.67, 0.83, 1, 1.33, and 2 g/L) was added to phosphate solutions (100 mg/L, pH = 7.0) to identify the optimum dosage. The centrifuge tubes were then put into a thermostatic water bath shaker at 298 K and shaken continuously for 12 h at 150 rpm. The supernatant was collected and filtered for the phosphate concentration analysis. The adsorption capacity and removal rate of phosphate were calculated using Equations (1) and (2), respectively.

$$Q_e={\displaystyle \frac{V}{\mathrm{M}}}\left(C_0-C_e\right)$$

(1)

$$R={\displaystyle \frac{C_0-C_e}{C_0}}\times 100\%$$

(2)

where Q_e (mg·g⁻¹) is the amount of phosphate adsorbed per gram of adsorbent, C_e (mg·L⁻¹) is the concentration of phosphate in the solution when the adsorption equilibrium is reached, C₀ (mg·L⁻¹) is the initial concentration of phosphate, M (mg) is the mass of the adsorbent, V (mL) is the volume of the solution, and R (%) is the removal rate of phosphate.

The effect of the initial pH (3.0–11.0) on the phosphate adsorption performance of Ca–Mg/BC was studied by constructing an adsorption system (dosage of 0.5 g/L, phosphate at a concentration of 100 mg/L). The initial pH was adjusted with a 1 mol/L NaOH and HCl solution, and the reaction was carried out for 12 h in a constant-temperature oscillator (298 K, 150 rpm). The remaining steps are the same as the experiment on the optimal dosage of biochar.

The effect of coexisting ions on the adsorption properties of Ca–Mg/BC phosphates was investigated by constructing an adsorption system. The specific experimental procedure was to add 15 mg of Ca–Mg/BC to a centrifuge tube containing 30 mL of a 100 mg/L phosphate solution and add NaCl, Na₂SO₄, Na₂CO₃, NaHCO₃, FeCl₃, and KCl to make the concentrations of Cl⁻, SO₄²⁻, CO₃²⁻, HCO₃⁻, Fe³⁺, K⁺, and Na⁺ reach 20, 40, 60, 80, and 100 mmol/L, respectively. The remaining steps are the same as the former.

The effect of Ca–Mg/BC on the actual biogas slurry in rural areas was also studied. The specific experimental procedure was to characterize the total phosphate content and initial pH of the actual biogas slurry. Subsequently, the gradient addition method (0.15–10 g·L⁻¹, dosage of 11 different gradients) was used to study the adsorption capability of the Ca–Mg/BC. The rest of the procedure is the same as the former. By calculating the equilibrium adsorption capacity and removal rate of Ca–Mg/BC, the optimal dosage in the actual biogas slurry treatment was finally determined.

In order to construct the adsorption isotherms, adsorption experiments were carried out with mixed solutions (0, 5, 10, 25, 50, 100, 200, 250, 400, 500, and 1000 mg·L⁻¹) at different concentrations. Experiments were carried out at 298, 308, and 318 K, and the initial pH of the solution was adjusted to 7 with HCl and NaOH. In order to construct the adsorption isotherms, adsorption experiments were carried out with mixed solutions (0, 5, 10, 25, 50, 100, 200, 250, 400, 500, and 1000 mg·L⁻¹) at different concentrations. Experiments were carried out at 298, 308, and 318 K, and the initial pH of the solution was adjusted to 7 with HCl and NaOH. The adsorption experiment was conducted at a 0.5 g/L adsorbent dosage with 12 h agitation at 150 rpm to reach equilibrium. After reaching equilibrium, the supernatant was collected and filtered through a 0.45 μm membrane filter for subsequent analysis. The adsorption data were fitted using the following Langmuir and Freundlich isotherm models:

$$Q_e=Q_{\max }{K}_L{C}_e/\left(1+K_L{C}_e\right)$$

(3)

$$Q_e=K_f{C}_e^{1/n}$$

(4)

where Q_max (mg·g⁻¹) represents the maximum adsorption capacity of the material, and K_f (mg²·n·g⁻¹·L⁻¹) and K_L (L·mg⁻¹) are the constants in the Freundlich and Langmuir models, respectively.

This study used nonlinear fitting methods to model and analyze adsorption isotherms and kinetic data. For the isothermal adsorption model, Langmuir and Freundlich nonlinear models were used to fit the adsorption isotherms to evaluate the adsorption mechanism and maximum adsorption capacity. For the adsorption kinetics model, pseudo-first-order and pseudo-second-order nonlinear kinetic models were used to analyze the adsorption rate process, and combined with the intraparticle diffusion model (linear fitting), the rate limiting steps of the adsorption process were further explored. All nonlinear fitting was based on the least squares method to optimize parameters, and the applicability of the model was evaluated through correlation coefficient (R²) and error analysis.

Adsorption kinetics were investigated. Specifically, a 100 mg·L⁻¹ phosphate solution was stirred at 298 K and 150 rpm. The pH of the solution was adjusted to 7. The concentration of the phosphate solution was measured by collecting samples at the designated time points (0, 5, 10, 15, 20, 30, 60, 120, 240, 360, and 480 min). The adsorption of phosphate by the prepared material was analyzed using the pseudo-first-order, pseudo-second-order, and intraparticle diffusion models. The formulas are as follows:

$$\ln \left(Q_e-Q_t\right)=\ln Q_e-k_{1}t$$

(5)

$$t/Q_e=1/\left(k_2{Q}_e^2\right)+t/Q_e$$

(6)

where k₁ (min⁻¹) and k₂ (g·mg⁻¹·min⁻¹) are the adsorption rate constants in the pseudo-first-order model and pseudo-second-order model, respectively.

$$Q_t=k_{id}{t}^{1/2}+C$$

(7)

where Q_t (mg·g⁻¹) is the amount of phosphate adsorbed by each gram of adsorbent at time t, k_id (mg·g⁻¹·min^−1/2) is the intra-particle diffusion rate constant, t (min) is the adsorption time, and C is a constant that depends on the width of the adsorption boundary layer.

Thermodynamic experiments were conducted at 298, 308, and 318 K to elucidate the natural occurrence and adsorption of phosphate. The Gibbs free energy (ΔG, kJ·mol⁻¹) was calculated using the experimental data obtained under various temperatures, employing the equation that incorporates changes in enthalpy (ΔH, kJ·mol⁻¹) and entropy (ΔS, kJ·mol⁻¹·K⁻¹). The corresponding equations are as follows:

$$\mathsf{\Delta }{G}_0=\mathsf{\Delta }{H}_0-T\mathsf{\Delta }{S}_0$$

(8)

$$\mathsf{\Delta }{G}_0=-RT\ln K_C$$

(9)

$$K_C=M_W\times 55.5\times 1000\times K_L$$

(10)

where R (8.314 J·mol⁻¹·K⁻¹) is the gas constant, K_C is the adsorption constant, and T (K) is the temperature. K_L is the Langmuir constant, which is the equilibrium constant for adsorption. M_w (g/mol) is the molecular weight of the adsorbate (g/mol) [16].

The potential of Ca–Mg/BC as a soil fertilizer was explored through pot experiments. The soil was obtained from the 0~20 cm surface soil of Southeast University (China).

Ca–Mg/BC saturated with phosphate solution (hereafter noticed as Ca–Mg/BC-P) was added to the soil at 5% of the soil dry weight. The environmental conditions of the pot experiment are shown in Table S1.

The leaching process was conducted using the methodology specified by the United States Environmental Protection Agency (USEPA 2014). Mung bean was used as the experimental plant. Commercial phosphate fertilizer, Ca–Mg/BC-P, and BC-P were mixed in 50 g of soil samples and added to the pot (groups A, B, and C respectively), and group D was the control group (no additives were added to the soil). Four green beans were planted in each pot and allowed to grow in a greenhouse at a temperature of about 298 K. Each grid was irrigated with 1.5 mL of tap water per day. Plants were selected for physical characterization on day 14. Equation (11) was used to calculate the germination rate (GP; %) of mung bean sprouts:

$$GP={\displaystyle \frac{n}{N}}\times 100\%$$

(11)

where n is the number of seeds germinated on the 14th day, and N is the total number of seeds.

All adsorption experiments were repeated three times, and the data were represented by the mean. The error bars represent the standard deviation of these repeated experiments.

2.3. Characterization and Analytical Method

AAn Mo-SS anti-spectrophotometer and a spectrophotometer (752 n spectrophotometer; Drawell, Chongqing, China) were used. The pH of the solutions was determined with a pH meter (pHS-3C, D-72G, HORIBA, Kyoto, Japan). The morphology and microstructure of the absorbent were analyzed via scanning electron microscopy (SEM; Regulus 8100, Hitachi, Tokyo, Japan). Elemental analysis of the surface was performed via energy dispersive X-ray spectroscopy (EDX) at the same sites where SEM was performed. N₂ adsorption–desorption isotherms were determined (ASAP 2460, Micromeritics, Norcross, GA, USA). Information on their S_BET, total pore volume, and average pore diameter (APD) parameters were obtained by testing the N₂ adsorption–desorption isotherms obtained at 77 K. The microstructure and crystal structure were analyzed via X-ray diffraction (XRD; D2 phaser, Bruker, Bremen, Germany). X-ray photoelectron spectroscopy (XPS; Scientific K-Alph, Thermo, Waltham, WA, USA) was performed using an Al K α source (hv = 1486.6 eV). The nickel IS10 catalyst was analyzed via Fourier transform infrared spectroscopy (FTIR) in the wave number range of 4000–400 cm⁻¹.

3. Results and Discussion

3.1. Structure–Activity Relationship of Phosphate Adsorption by Ca–Mg/BC

SEM analysis revealed the influence of thermochemical modification on the morphological characteristics and composition properties of biogas-residue-based biochar. A comparison of the microscopic morphology of the BR and BC samples at 5 μm in Figure 1a,b, shows that the pristine BC has a relatively dense surface structure. After chemical impregnation combined with a thermal modification process, the Ca–Mg/BC surface formed an irregular layered stacking structure. Figure 1c demonstrates that modified surface particles showed pronounced aggregation, resulting in substantially enhanced surface roughness [17].

In order to further analyze the elemental distribution characteristics, the elemental imaging characterization of Ca–Mg/BC was carried out via SEM-EDX analysis (Figure 1d–i). The spectra indicated the spatially overlapping distribution of Mg, Ca, C, and O, with Mg and Ca mass fractions reaching 18.6% and 18.2%, respectively. Thess findings confirmed that the chemical impregnation facilitated the targeted deposition of Ca and Mg minerals on the surface of the biochar matrix under the thermal modification condition of 700 °C, yielding inorganic–organic composite structures with a characteristic morphology [18].

Information on the parameters of S_BET, total pore volume, and average pore diameter (APD) of BR and BC is presented (

Table 1. Comparison of S_BET, total pore volume, and APD for BR, BC, and Ca–Mg/BC.

Parameters	Adsorbents
	BR	BC	Ca-Mg/BC
SBET (m2·g−1)	0.20	7.97	72.33
Total pore volume (cm3·g−1)	0.0012	0.061	0.12
APD (nm)	23.66	3.62	25.28

). Ca–Mg/BC was obtained by analyzing its adsorption–desorption isotherms obtained at 77 K (Figure 2a). The S_BET values of the three materials were 0.2, 7.97, and 72.33 cm²/g, respectively. This order of magnitude increase indicated that the metal modification process forms richer microporous structures on the material surface and the dispersion of Ca–Mg-containing oxides on the carrier surface leads to the expansion of layered structures. Meanwhile, the layered structure loading further enhanced pore expansion, potentially improving the phosphate-adsorption capacity of the Ca–Mg/BC. All materials exhibited average pore diameters (APDs) ranging from 2 to 50 nm, confirming their classification as typical mesoporous materials [19]. Among them, for BC and Ca Mg/BC, the APDs were 3.62 and 25.28 nm, respectively. As shown in Figure S1, Ca–Mg/BC exhibits the minimum average pore diameter corresponding to the highest surface area and pore volume, indicating that it may have excellent phosphate-removal performance [20]. Consistent with the IUPAC classification, BR, BC, and Ca–Mg/BC displayed type IV adsorption isotherms, with Ca–Mg/BC showing an H₃-type hysteresis loop characteristic of an irregular and porous structural feature [21]. It was noteworthy that the three materials reached adsorption limits within the relative pressure range (P/P₀) of 0.9 to 1.0, suggesting that the biochar might be a mesoporous, hierarchical porous structure with a broad pore size distribution and high pore connectivity. This structural features are likely advantageous for the phosphate adsorption.

The chemical bond structures of BR, BC, and Ca–Mg/BC were characterized using FTIR spectroscopy. As is shown in Figure 2b, the characteristic peaks at 3440 cm⁻¹ and 1626 cm⁻¹ are associated with hydroxyl groups (O-H bending vibrations from either adsorbed water or hydroxyl groups). The peak intensity showed a pattern of BR > Ca-Mg/BC > BC, which could be attributed to the partial elimination of surface-adsorbed water and hydroxyl groups during BC calcination. In contrast, Ca–Mg/BC retained residual O-H groups due to the generation of metal hydroxides as Mg(OH)₂. Meanwhile, the hygroscopicity of the metal oxides partially restored the peak intensity. The medium-intensity peak at 3640 cm⁻¹ was assignable to the O-H stretching vibration of Ca(OH)₂, whereas the high-intensity peak at 3700–3650 cm⁻¹ corresponded to the O-H vibration of Mg(OH)₂. In the 1060 cm⁻¹ region; the Si-O-Si peaks in BR exhibited stronger intensity compared to BC and Ca-Mg/BC, suggesting that calcination facilitated the transformation of the Si into a stable SiO₂ crystal structure. Meanwhile, the metal oxides in Ca–Mg/BC might interact with the C-O groups through the formation of carbonates or chelates, and the loading of Ca and Mg could obscure Si-O groups, collectively resulting in reduced peak intensity. Characterization of the low-frequency region showed that the vibrational peaks around 450 cm⁻¹ originated from the Ca-O bond, while the range 500–600 cm⁻¹ contained the Mg-O vibrations. The 800–850 cm⁻¹ range showed O-H vibrations specific to Ca(OH)₂. These findings demonstrated that Ca-Mg/BC possessed abundant surface functional groups, such as O-H, metal-oxygen, and metal-hydroxide bonds, which might enhance phosphate adsorption [22,23].

The evolution of the surface chemical composition of BR, BC, and Ca–Mg/BC was systematically characterized via XRD analysis. As is shown in Figure 2c, compared to BR and BC, Ca–Mg/BC showed characteristic diffraction peaks at 2θ = 36.8° (200), 42.82° (220), 62.17° (222), and 78.44° (400) that matched the standard card of PDF#71-1176 for MgO. The characteristic peaks of the CaO crystalline phase conforming to PDF#37-1497 were also detected at 32.19° (111), 37.34° (200), 53.84° (220), 64.13° (311), 67.36° (222), and 79.63° (400). This confirmed that chemical impregnation combined with thermal modification at 700 °C could lead to the pyrolytic transformation of Ca and Mg and the generation of MgO and CaO crystalline phases on the biochar surface.

The elemental composition of BR, BC, and Ca–Mg/BC materials was characterized via XPS. The energy spectrum analysis showed (Figure 2d and

Table 2. The elemental composition (atomic%) measured via XPS.

Adsorbent	C	O	Ca	Mg	Fe	Si	N	P	Cl
BR	65.68	15.21	1.54	0.99	2.78	1.86	4.14	2.37	5.43
BC	56.67	22.08	2.05	1.21	2.1	2.9	5.49	1.9	5.6
Ca-Mg/BC	24.68	41.46	16.44	8.76	1.24	0.67	2.95	0.78	3.02

) that Ca-Mg/BC is mainly composed of C (24.68%), O (41.46%), Ca (16.44%), and Mg (8.76%) and also contains trace amounts of Si (0.67%) and Fe (1.24%). The total content of Ca and Mg elements enriched in the material reached 25.2%, suggesting that there may be a metal oxide/hydroxide loading structure on the surface, and this characteristic modification could substantially enhance the ligand-adsorption capacity of the material for phosphate in water. Further analysis revealed that the Fe element (2.78%) was inherent in the pristine BR. The Fe content in Ca–Mg/BC was 1.24%, and the biogas residue on the surface contained Fe components. Under the pH conditions of the solution, the surface of iron-containing oxides was positively charged, which facilitated the co-precipitation of negatively charged phosphate. Fe could form insoluble precipitates with phosphate (such as FePO₄). This multi-mechanism synergistic effect was to promote the phosphate-adsorption capacity of Ca–Mg/BC [24].

3.2. Optimization of Synthesis Conditions for Sustainable Biochar Production

The metal salt solution impregnation concentration and calcination temperature critically influence the phosphate recovery efficiency of biochar-residue-based biochar. Therefore, the optimization of the preparation conditions could improve the capacity of phosphate adsorption. The following reaction system was built in order to examine how various impregnation concentrations affected the modified biogas-residue-based biochar’s adsorption efficiency The concentrations of Mg(CH₃COO)₂ and Ca(CH₃COO)₂ were equal, both set to 0.1, 0.2, 0.3, 0.4, 0.5, and 1 mol/L. The samples were named Ca–Mg/BC-0.1, Ca–Mg/BC-0.2, Ca–Mg/BC-0.3, Ca–Mg/BC-0.4, and Ca–Mg/BC-0.5, respectively. As can be seen from the Figure 3a, the adsorbent modified through impregnation with different concentrations showed different adsorption capacities. From 35.03 mg/g in the unimpregnated modification to 156.08 mg/g in the impregnation modification at a concentration of 0.3 mol/L, the adsorption property progressively increased as the concentration of the metal salt solution increased The adsorption amount was kept stable with the continued increase of the impregnation concentration. The XRD analysis in Figure 3b reveals that as the Mg(CH₃COO)₂ and Ca(CH₃COO)₂ concentrations increase, the characteristic peaks of CaO and MgO in Ca–Mg/BC progressively intensify before stabilizing. This trend indicated that a concentration of 0.3 mol/L for both precursors maximizes the loading of MgO and CaO onto the material surface. By balancing the economic feasibility with adsorption performance, the optimal concentrations of Mg(CH₃COO)₂ and Ca(CH₃COO)₂ for subsequent experiments were determined to be 0.3 mol/L.

In order to examine the impact of the calcination temperature on phosphate adsorption, biogas-residue-based biochar was modified at various temperatures while maintaining an initial phosphate concentration of 100 mg/L, a shaking rate of 150 rpm, and an adsorbent dosage of 0.5 g/L. The results are shown in Figure 3c. It could be seen that the adsorption capacity of phosphate by biochar increased gradually with the increase in temperature and remained stable after 700 °C. This is probably because when the temperature was not high enough, the pore expansion of the biochar was not obvious enough, resulting in a small increase in the specific surface area of the biochar, and some of the adsorption sites were still buried inside the biochar. And when the calcination temperature was too high, it would lead to the collapse of the pore on the surface of the biochar, which would also lead to the above problems. The calcination temperature of 700 °C was chosen for the subsequent experiments.

3.3. Adsorption Isotherm, Kinetics, and Thermodynamics

Through a systematic investigation of phosphate adsorption isotherms, kinetics, and thermodynamics combined with a quantitative characterization of the adsorption capacity, reaction kinetics, and energy variations, this study elucidated the synergistic mechanism between surface complexation and precipitation at the Ca–Mg/BC solid–liquid interface.

The isothermal adsorption modeling systematically revealed the difference in the adsorption mechanism of phosphate between Ca–Mg/BC and BC (Figure 4a,b). Freundlich and Langmuir model fitting results demonstrated that the adsorption properties of Ca–Mg/BC were more consistent with the Freundlich model (R² = 0.9757), demonstrating that the adsorption process occurred as a multimolecular layer with an uneven surface energy distribution and a complex mechanism of action [25]. As is shown in

Table 3. Parameters of adsorption isotherm models for phosphate adsorption by Ca–Mg/BC and BC at 298 K.

Model	Parameter	Ca-Mg/BC	BC
Langmuir	qe (mg·g−1)	1596.59	226.81
	KL (L·mg−1)	0.0029	0.0029
	R2	0.9683	0.9875
Freundlich	KF	23.17	4.50
	1/n	0.5879	0.5370
	R2	0.9757	0.9443

, the model parameter 1/n = 0.5870 (0 < 1/n < 1) further confirms that the material is characterized by desirable site heterogeneity, which is conducive to the adsorption process spontaneously [26]. In contrast, the adsorption behavior of BC was consistent with the Langmuir monomolecular layer adsorption pattern, indicating a strong homogeneity of its surface-active sites. This shift in the adsorption mechanism might stem from the evolution of surface functional group reconfiguration brought about by bimetallic modification.

The fitted curves of the adsorption kinetics of Ca–Mg/BC and BC are shown in Figure 4c,d, the adsorption capacity of phosphate by Ca–Mg/BC increases rapidly within 1 h and reaches equilibrium after 1 h. A related study found that the Ca–Mg/BC equilibration time was quicker than that of other biochars. The strong affinity between the negatively charged phosphate and the positively charged surface of Ca–Mg/BC may be the cause of the fast phosphate absorption by Ca–Mg/BC [27]. The adsorption capacity of phosphate by BC increased gradually, and the rate slowed down after 6 h. The pseudo-first-order and pseudo-second-order kinetic models fit the phosphate-adsorption capacity process by Ca–Mg/BC, and the fitting outcomes are displayed in

Table 4. Ca–Mg/BC adsorption kinetic parameters of phosphate.

Model	Parameter	Ca-Mg/BC	BC
Pseudo-first-order	qe (mg·g−1)	154.14	36.22
	k1 (min−1)	0.41	0.01
	R2	0.9778	0.9837
Pseudo-second-order kinetic	qe (mg·g−1)	158.87	44.63
	k2 (g·mg−1·min−1)	0.00718	2.36 × 10−4
	R2	0.9858	0.9869
Intraparticle diffusion	ki1 (mg·g−1·min−0.5)	31.49	0.2021
	R2	0.8565	0.8869
	ki2 (mg·g−1·min−0.5)	0.69	0.01546
	R2	0.8045	0.9817

. The simulation of Ca–Mg/BC using the pseudo-second-order kinetic model had a higher R² (0.9858) than the pseudo-first-order kinetic model (0.9778), suggesting that the pseudo-second-order kinetic model could more accurately describe the phosphate-adsorption process by Ca–Mg/BC. The fitting results indicated that the process of the adsorption of phosphate was mainly controlled by intermolecular interactions, electron exchange between the adsorbent and adsorbate, and the formation of new compounds, which might react with phosphate on the surface of Ca–Mg/BC to produce new compounds or generate electrostatic forces [28]. The R² of the pseudo-first-order kinetic model fit for phosphate adsorption by BC (R² = 0.9869) was higher than that of the pseudo-second-order model (R² = 0.9837), suggesting that the adsorption capacity of phosphate by BC was mainly controlled by physical effects.

To further study the mechanism of phosphate adsorption by Ca–Mg/BC and BC, the kinetic experimental data were fitted and analyzed using the intraparticle diffusion model. As shown in Figure 4e, the nonlinear relationship between q_e and t^1/2 (not passing through the origin) indicates that the phosphate adsorption on Ca–Mg/BC involves multiple successive steps. The adsorption data were described by a two-phase linear model. In two segments of the intra-particle diffusion process, k_i1 (31.49) > k_i2 (0.69), and the boundary layer thickness 0 < C₁ < C₂, this difference suggested that the overall adsorption rate was affected by both intra-particle diffusion and external diffusion, but it was mainly controlled by the external diffusion. The adsorption rate decreased over time as the diffusion resistance until equilibrium was reached, with the initial adsorption rate exerting greater control over the overall process kinetics. For BC, as is shown in Figure 4f, k_i1 (0.20) > k_i2 (0.015), indicating that the adsorption capacity of phosphate was all determined by external diffusion.

The Langmuir model fits the phosphate-adsorption process by Ca–Mg/BC at 298 K, 308 K, and 318 K, as illustrated in Figure 5. The adsorption capacity of 100 mg/L of phosphate by the Ca–Mg/BC rose from 158.39 mg/g to 169.87 mg/g when the temperature was raised from 298 K to 318 K. Thermodynamic calculations were performed on the data of phosphate adsorption by Ca–Mg/BC at three temperatures, and the results are shown in

Table 5. Thermodynamic parameters of adsorption of phosphate by Ca–Mg/BC.

Temperature (K)	KC	ΔH (kJ·mol−1)	ΔS (kJ·mol−1·K−1)	ΔG (kJ·mol−1)
298	15,145.71	2.48	0.072	−23.85
308	20,317.41			−24.58
318	27,019.52			−25.28

. ΔH > 0 demonstrated that the phosphate adsorption process was heat-absorbing, and the adsorption mechanism might be chemisorption, such as surface complexation or ion exchange. It might be that increasing the temperature would increase the surface active sites of the adsorbent or increase the phosphate diffusion rate. ΔG was negative, and its magnitude decreased with the increasing temperature, indicating that the adsorption process was spontaneous, but the spontaneity decreased with an increasing temperature. ΔS> 0 indicated an increase in disorder at the interface between Ca–Mg/BC and phosphate, which was consistent with phosphate adsorption on solid surfaces.

3.4. Analysis of Factors Affecting Phosphate Recovery by Biochar

3.4.1. Optimizing the Biochar Dosage for Sustainable Phosphate Recovery

As is shown in Figure 6a, in the experiments of the phosphate-adsorption capacity by Ca–Mg/BC, as the adsorbent dosage was gradually increased from 0.17 g/L to 2 g/L, the adsorption amount showed a substantial decreasing trend, going from 361.27 mg/g to 144.81 mg/g. Nevertheless, the removal rate of phosphate was substantially increased from the initial 60.21% to 97% and stabilized. In terms of the adsorption mechanism and the dynamic equilibrium of the system, with the increase in the dosage, although the adsorption capacity per unit mass of Ca–Mg/BC decreased due to the under-utilization of active sites or inter-particle aggregation effect, the number of total adsorption sites in the system increased substantially, which improved the overall removal capacity of phosphate [29]. The removal rate was close to saturation (97%) when the dosage reached 0.5 g/L, and although continuing to increase the dosage could slightly improve the removal effect, the excessive dosage of adsorbent meant a waste of resources. Therefore, considering the principles of the adsorption efficiency, adsorption amount, and economy, 0.5 g/L was finally selected as the optimal dosage. At this concentration, a high removal rate of about 95% could be achieved, and a reasonable balance between the adsorbent dosage and adsorption efficiency could be achieved, which provided an economically feasible optimization scheme for practical applications.

3.4.2. Effect of pH on Phosphate Adsorption

The form of phosphate in solution was affected by the pH, which affects its adsorption on Ca–Mg/BC [30]. As is shown in Figure 6b, the optimum pH range for phosphate adsorption is 7.0–9.0, with adsorption reaching 134.83 mg/g. Although adsorption was unfavorable at pH < 5 or pH > 9, high adsorption capacity was still observed across a pH range of 5.0–13.0. The probable reason was the different forms of phosphates at the different pHs. When the pH was 7–9, phosphate exists mainly as HPO₄²⁻ and some PO₄³⁻. The adsorbent Ca–Mg/BC, on the other hand, contains Ca²⁺ and Mg²⁺, which might form precipitates like CaHPO₄, Ca₃(PO₄)₂, and Mg₃(PO₄)₂ through electrostatic attraction of the metal cations to the phosphate. At pH = 7–9, the hydroxylation of these metal ions might facilitate binding to phosphate and the formation of surface precipitates or coordination complexes. At pH < 5, H₂PO₄⁻ dominated, and high H⁺ concentrations competing with metal ions for adsorption sites. Protonation of the adsorbent surface might also reduce electrostatic attraction to phosphate ions. In addition, metal ion leaching could decrease active sites, lowering the adsorption capacity. At pH > 9, phosphate existed mainly in the form of PO₄³⁻. The surface of the adsorbent might be negatively charged as the deprotonation of hydroxyl groups on the surface of metal oxides at pH > 9, resulting in electrostatic repulsion, which was unfavorable for adsorption. However, the experimental result showed a high adsorption capacity of Ca–Mg/BC of phosphates even at pH > 9, which might indicate the existence of other mechanisms. (1) The formation of metal hydroxides like Ca(OH)₂ and Mg(OH)₂, which were able to immobilize phosphates via surface adsorption or precipitation reactions [31]. In addition, it might promote the formation of more stable complexes or precipitates of metals with phosphates when the pH > 9, such as the formation of Ca₅(PO₄)₃(OH) and Mg₃(PO₄)₂, thus maintaining a high adsorption capacity of Ca–Mg/BC [32,33]. (2) At pH > 9, phosphate was mainly composed of HPO₄²⁻ (pH = 9~12), which could still bind to Ca/Mg through ionic or chemical bonds. Only when the pH > 12.3, PO₄³⁻ became the dominant form, and the adsorption capacity might substantially decrease [34]. The synergistic effect of the metal loading of Ca–Mg/BC and biochar matrix endowed it with broad-spectrum pH adaptability, especially in alkaline conditions to maintain efficient adsorption through precipitation, breaking through the pH limitation of conventional adsorbents.

3.4.3. Effect of Coexisting Ions on Phosphate Adsorption

In an aqueous environment, coexisting ions (HCO₃⁻, CO₃²⁻, SO₄²⁻, Cl⁻, Na⁺, K⁺, and Fe³⁺) constitute key factors influencing phosphate adsorption by Ca–Mg/BC. The experimental results show that the strength of the effect of the four coexisting anions on the adsorption properties is in the order CO₃²⁻ > HCO₃⁻ > SO₄²⁻ > Cl⁻, and the strength of the effect of cations is in the order Fe³⁺ > Na⁺ > K⁺ (Figure 6c,d). The inhibition of phosphate adsorption by CO₃²⁻, HCO₃⁻, SO₄²⁻, and Fe³⁺ exhibited the concentration-dependent inhibition of phosphate adsorption: as concentrations increased from 20 to 100 mmol/L, the phosphate-adsorption capacity decreased to 74.91, 94.30, 127.03, 2.02, and 41.13 mg/g, respectively. In contrast, Na⁺, K⁺, and Cl⁻ showed a facilitating effect, with adsorption increasing to 182.63, 179.67, and 182.67 mg/g, respectively. The above phenomenon may stem from the following mechanisms: (1) ligand complexation occurs as HCO₃⁻ and CO₃²⁻ form stable complexes (e.g., CaCO₃) with metal sites on Ca–Mg/BC, thereby blocking active site closure [35,36,37]. (2) The electrostatic effect manifests when Fe³⁺ increased the adsorbent’s positive charge density through electrostatic attraction, promoting phosphate co-precipitation. However, at elevated concentrations, Fe³⁺ hydrolyzes to form colloidal particles that occlude active sites [38]. (3) The ionic strength effect explains how low-charge-density cations (Na⁺, K⁺) compress the electric double layer, reducing electrostatic repulsion between the phosphate and adsorbent [39]. However, it is noteworthy that these co-existing ions did not substantially affect the phosphate-adsorption capacity of Ca–Mg/BC in the conventional concentration range of the actual wastewater.

3.4.4. Sustainable Phosphorus Recovery from Actual Biogas Slurry Using Ca–Mg/BC

The resource recovery potential of Ca–Mg/BC was further evaluated using an actual biogas slurry (TP = 15.51 mg/L, pH = 8.64 ± 0.01). As shown in Figure 7, increasing the dosage from 0.15 g/L to 8 g/L enhanced phosphate removal from 14.92% to 85.23%, reducing residual concentrations to 4.4 mg/L and 0.55 mg/L, respectively. Compared to simulated wastewater, the removal efficiency decreased in the biogas slurry, likely due to competitive adsorption from organic acids, metal ions, and other matrix components that occupy active sites (e.g., Ca/Mg-OH, metal oxides [40]. Therefore, in the actual application of the project, this is according to the water quality characteristics of the optimization of operating parameters, through the pre-experiment to determine the minimum dosage in line with the emission standards such as TP ≤ 0.5 mg/L, in order to balance the effect of treatment and operational sustainability.

3.5. Phosphate Adsorption Mechanism Investigation

To elucidate the interfacial chemical behavior of Ca–Mg/BC during phosphate adsorption, systematic FTIR analyses of Ca–Mg/BC, Ca–Mg/BC-P, and Des-Ca–Mg/BC-P were conducted (Figure 8a). The appearance of the characteristic peak at 3697 cm⁻¹ suggested that Mg/Ca-based hydroxides like Mg(OH)₂ and Ca(OH)₂ were generated on the surface of Ca–Mg/BC-P after phosphate adsorption. The O-H stretching vibrations were directly related to the hydration of metal oxides or the formation of phosphate–metal complexes. The retention of this peak after desorption confirmed that the hydroxide structure remained stable during phosphate immobilization [41]. The intensified peak at 1060 cm⁻¹ in Ca–Mg/BC-P likely results from the combined effects of phosphate adsorption-induced metal overlayer rupture (exposing Si-O groups) and the P-O stretching vibration of PO₄³⁻ (1050–1100 cm⁻¹). And the intensity of the peaks weakened but did not disappear after desorption, suggesting that part of the phosphate was stabilized on the material surface through chemical bonding [42]. The characteristic peak at 566 cm⁻¹ in Ca–Mg/BC represents the P-O bond and the stretching vibration of the Ca/Mg-O-P bond in metal phosphates (Ca/Mg-PO₄), confirming phosphate chemisorption with Ca²⁺/Mg²⁺ to form stable coordination compounds [43]. The persistence of this peak after desorption further confirmed that the mechanism of phosphate immobilization involved chemical bonding and crystalline-phase precipitation rather than merely physical adsorption or ion exchange. The attenuation of the Ca/Mg-O vibrational peak near 500 cm⁻¹ post-adsorption indicated that the Ca²⁺/Mg²⁺ active sites on the surface of the material were involved in the phosphate coordination reaction and were converted to the metal–phosphate phase [27,44,45]. In conclusion, these findings demonstrated that phosphate adsorption by Ca–Mg/BC might depend on the chemical transformation of Ca/Mg oxides and hydroxides on its surface, which ultimately produced stable metal phosphates (Ca/Mg-PO₄).

An XRD analysis was performed on Ca–Mg/BC, Ca–Mg/BC-P, and Des-Ca–Mg/BC-P to investigate their phase composition (Figure 8b). The characteristic peaks of MgO and CaO of Ca–Mg/BC were weakened after phosphate adsorption, while the diffraction peaks of Mg(OH)₂ were generated probably due to the dissolution of MgO in water. And the appearance of characteristic peaks corresponding to Ca₅(PO₄)₃(OH) and Mg₃(PO₄)₂ provided direct evidence for the formation of Ca/Mg-PO₄ complexes [46,47].

XPS characterization was employed to further elucidate the phosphate-adsorption mechanism. As shown in Figure 8c, the phosphate content in Ca–Mg/BC-P increased, as evidenced by the distinct P 2p peak at 133.48 eV. The results of the characteristic peak fits (133.42 eV and 133.97 eV) could be the chemical state of phosphate in Mg₃(PO₄)₂ and Ca₅(PO₄)₃(OH) [34], which directly confirmed the chemical fixation of phosphate [48,49,50].

The binding energy analysis showed that the Mg 1s peak is shifted to high binding energy by 0.1 eV (1303.48 → 1303.58 eV), suggesting that the Mg active site is involved in the phosphate coordination (Figure 9a). The Ca 2p main peak exhibited a more pronounced shift of 0.71 eV (349.78 → 350.49 eV), confirming the conversion of Ca from the initial CaO/Ca(OH)₂ to the calcium phosphate compounds (Figure 9b). In addition, the complete absence of the P 2p peak in pre-adsorption samples ruled out the possibility of physical adsorption (Figure 9c,d). Synthesis showed that Ca–Mg/BC generated Ca/Mg-PO₄ crystalline phases through hydrolysis, electrostatic attraction, and chemical precipitation of surface metal active sites (Ca²⁺/Mg²⁺) for efficient phosphate immobilization. The following reactions occur (M for Mg and Ca):

$${\mathrm{MO}+\mathrm{H}}_2\mathrm{O}\to \mathrm{M}{\left(\mathrm{OH}\right)}_2$$

(12)

$${\mathrm{MO}+\mathrm{H}}_2\mathrm{O}\to {\mathrm{MOH}}^{+}{+\mathrm{OH}}^{-}$$

(13)

$${\mathrm{M}}^{2+}{+2\mathrm{H}}_2\mathrm{O}\to \mathrm{M}{\left(\mathrm{OH}\right)}_2{+2\mathrm{H}}^{+}$$

(14)

$$3{\mathrm{M}}^{2+}{+2\mathrm{PO}}_4^{3-}\to {\mathrm{M}}_3{\left({\mathrm{PO}}_4\right)}_2\downarrow$$

(15)

For Ca²⁺, further hydroxyapatite (Ca₅(PO₄)₃(OH)) will be generated:

$${5\mathrm{Ca}}^{2+}+{3\mathrm{PO}}_4^{3-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}_5{\left({\mathrm{PO}}_4\right)}_3\left(\mathrm{OH}\right){+\mathrm{H}}^{+}$$

(16)

3.6. Sustainable Crop Growth Validation Test

This study systematically investigated the sustainable agronomic potential of Ca–Mg/BC as a soil amendment using a controlled potting system. The experimental data are shown in

Table 6. Results of pot experiment.

Group	Stem Length (cm)	Root Length (cm)	Average Wet Weight (g)	Average Dry Weight (g)	Phosphorus (g/kg)
A (Commercial phosphate fertilizer)	17.44	3.52	0.65	0.084	3.86
B (Ca-Mg/BC-P)	15.98	3.24	0.59	0.069	3.79
C (BC-P)	11.08	3.10	0.39	0.055	2.13
D (No additives)	10.13	3.18	0.39	0.058	2.01

. The initial germination rate of mung bean seeds in the four treatment groups showed differences of 80%, 73%, 53%, and 60%, respectively, which indicated that Ca–Mg/BC-P did not substantially inhibit seed viability. The commercial phosphate fertilizer treatment exhibited a 12–20% higher seeding rate compared to the other groups.

The morphological development of mung bean sprouts on the first day is shown in Figure 10a. The experimental results show that by the 14th day (Figure 10b,c), the plants in groups A and B applied with commercial phosphate fertilizer and Ca–Mg/BC-P exhibited a substantial growth advantage. The root and stem growth indexes of these two groups were about 15% higher than those of group C, with BC-P addition, and group D, with blank soil. The results of biomass measurements are shown in Figure 10d, where the fresh weights of the plants reached 0.65 g (84 mg dry weight) in group A and 0.59 g (69 mg dry weight) in group B. This result is substantially higher than 0.39 g (55–58 mg dry weight) in groups C and D. Further testing of plant tissues confirmed that mung bean sprouts in groups A and B contained 3.86 g/kg and 3.76 g/kg of phosphate, respectively, representing an 80% increase over the control group (2.01–2.13 g/kg).

Although the growth-promotion effect of Ca–Mg/BC was 5–10% lower than that of commercial phosphate fertilizer, its overall efficacy remained substantially superior to the conventional treatment group. This advantage might stem from the material’s unique pore structure, which can improve soil sustainability through the gradual release of Mg²⁺/Ca²⁺ and enhance water and fertilizer retention owing to its high specific surface area. Thus, phosphate-saturated Ca–Mg/BC demonstrates potential as an excellent soil amendment, offering a resource-efficient strategy for improving soil fertility and nutrient-cycling in agricultural systems.

4. Conclusions

In this study, a bimetallic-loaded modified biogas residue-based biochar was prepared from industrial iron-containing anaerobic digestate. Ca–Mg/BC with a high adsorption property was successfully prepared using the optimized process of Ca(CHCOO)₂ and Mg(CHCOO)₂ mixed salt solution impregnation combined with 700 °C calcination. Compared with BC, the specific surface area of the modified material was enhanced to 67.65 m²/g, and the phosphate-adsorption capacity was substantially improved.

(1)

Circular wastewater treatment potential was evidenced by 92.56% phosphate removal from a 100 mg/L solution and >80% removal from an actual biogas slurry (8 g/L dosage, 15.51 mg/L).

(2)

The multilayer adsorption properties were confirmed through a kinetic analysis, which revealed that the adsorption process followed a pseudo-second-order kinetic model (R² = 0.9858). The isothermal adsorption data also showed a high degree of agreement with the Freundlich model; 1596.59 mg/g was the maximum adsorption capacity determined using the Langmuir model at 298 K, which is 7.04 times higher than that of BC The thermodynamic parameters ΔG < 0 and ΔH = +1.63 kJ/mol confirmed that the process was a spontaneous heat absorption reaction.

(3)

Mechanistic characterization revealed a dual-pathway sustainable phosphorus-management system. Surface-hydrolyzed Ca²⁺ and Mg²⁺ initially capture phosphate through electrostatic attraction, forming Ca₅(PO₄)₃(OH) and Mg₃(PO₄)₂ precipitates for long-term nutrient recycling. The results of the pot experiment demonstrated circular economy benefits. The phosphate-saturated material as a soil conditioner could increase the stem length, root length by 15%, and fresh weight by 51%, confirming that it combines the functions of nutrient slow-release and a soil conditioner.

(4)

This study established a sustainable technical pathway of “Waste Recycling-Pollution Control-Agricultural Application”, offering an eco-friendly solution for biogas residue and slurry utilization. The bimetallic modification strategy significantly improved the phosphate-adsorption capacity of biochar, demonstrating great potential for sustainable phosphate removal from wastewater and nutrient recycling in agriculture.