Effects of Total Calcium and Iron(II) Concentrations on Heterogeneous Nucleation and Crystal Growth of Struvite

Abstract

This study investigated the effects of calcium (Ca²⁺) and iron (II) Fe²⁺ concentrations (0–500 mg/L) on the heterogeneous nucleation and crystallization behavior of struvite (MgNH₄PO₄·6H₂O) through controlled batch precipitation experiments. Struvite formed under different Ca²⁺ and Fe²⁺ concentrations were systematically characterized using XRD, SEM, FTIR, and XPS, while real-time pH and redox potential (Eh) monitoring was employed to elucidate reaction dynamics and thermodynamic speciation and saturation indices were calculated, and classical nucleation theory (CNT) was applied to interpret nucleation behavior. The results show that Ca²⁺ primarily suppresses struvite formation through bulk-phase competition with Mg²⁺ for phosphate, diverting phosphate into Ca–P phases and progressively reducing struvite supersaturation, which leads to decreased crystallinity and distorted Crystal habit. In contrast, Fe²⁺ does not form detectable crystalline Fe-P phases but inhibits struvite crystallization mainly through surface-mediated processes. Surface analyses indicate that Fe-bearing species adsorb onto struvite surfaces and promote amorphous Fe-P deposition, increasing interfacial resistance to nucleation and growth. CNT analysis further reveals that Ca²⁺ inhibition is governed by reduced thermodynamic driving force, whereas Fe²⁺ inhibition is dominated by surface-related kinetic barriers. This study provides mechanistic insight into ion-specific interference during struvite crystallization and offers guidance for optimizing phosphorus recovery in ion-rich wastewater systems.

1. Introduction

Phosphorus is an essential and non-substitutable element for global food production, yet its primary source, phosphate rock, is non-renewable and increasingly scarce. With the continued depletion of high-grade mineral reserves and the growing dependence on imports in many regions, recovering phosphorus from secondary resources has become a strategic pathway to enhance resource security [1,2,3]. Municipal and agricultural wastewaters contain substantial amounts of reactive phosphorus, offering promising opportunities aligned with circular-economy and sustainability objectives. Among various recovery technologies, struvite (MgNH₄PO₄·6H₂O) crystallization has attracted considerable attention because it simultaneously removes phosphorus, mitigates scaling in wastewater treatment systems, and produces a slow-release fertilizer with commercial value [4,5,6]. Efficient and controllable struvite formation is therefore important for advancing resource-oriented wastewater treatment [7].

However, the crystallization behavior of struvite is strongly influenced by coexisting ions in wastewater, particularly multivalent cations such as Ca²⁺ and Fe²⁺, which can significantly affect nucleation, crystal growth, and final product purity [8,9]. Calcium often competes with magnesium for phosphate and may induce the formation of calcium phosphates, thereby altering supersaturation conditions and shifting crystallization pathways [10]. Iron, commonly originating from chemical phosphorus-removal processes or industrial effluents, may form Fe-phosphate precipitates or adsorb onto crystal surfaces, influencing both struvite yield and habit [11]. These ion-specific effects complicate the regulation of crystallization processes and can lead to considerable variability in product quality in real wastewater systems [12].

Although the inhibitory roles of Ca²⁺ and iron have been reported, several mechanistic gaps remain [13,14]. For Ca²⁺, many studies emphasize thermodynamic competition and the lower solubility of Ca-P phases compared with struvite under alkaline conditions. However, the concentration-dependent behavior of Ca²⁺ interference is not always linear: the balance between Ca-P co-precipitation and struvite crystallization may shift with ionic strength, pH buffering, and nucleation regime (homogeneous vs. heterogeneous). In particular, under heterogeneous nucleation conditions, which are relevant to engineered crystallizers where surfaces or seed materials are present, how Ca²⁺ modulates nucleation kinetics, critical nucleus formation, and crystal development has not been sufficiently clarified across broad concentration gradients. Notably, dairy and livestock wastewater streams are typically richer in calcium than magnesium, and Ca²⁺ competes with Mg²⁺ for phosphate binding at neutral to alkaline pH, leading to the formation of either crystalline or amorphous calcium phosphates that can interfere with struvite precipitation [15]. For iron, prior research has predominantly focused on Fe³⁺ because it is commonly associated with oxidative chemical phosphorus removal and hydrolysis-driven precipitation, whereas Fe²⁺, which is the prevailing iron species in sludge digestate and livestock wastewater under reducing conditions, has received far less investigation [16,17]. The surface-mediated Fe²⁺ inhibition mechanism is particularly relevant in anaerobic digestion and livestock wastewater treatment scenarios, where reducing conditions maintain iron predominantly in the ferrous state, contrasting with the Fe³⁺-dominated chemistry observed in aerobic environments [18]. Since Fe²⁺ is widely present in these matrices at measurable levels [19], and the behavior of Fe²⁺ can differ fundamentally from Fe³⁺, not only in redox chemistry but also in how it interacts with phosphate and crystal surfaces. Fe²⁺ may form ferrous phosphate under favorable conditions, yet it may also exert strong interfacial effects via adsorption or surface complexation, potentially altering nucleation barriers and growth-unit integration without necessarily producing detectable crystalline secondary phases. As a result, the interference mechanism of Fe²⁺ in struvite systems remains insufficiently understood, particularly under well-controlled anaerobic conditions. Clarifying its effect on struvite nucleation, growth, and phase composition remains essential. More importantly, Ca²⁺ and Fe²⁺ are often collectively regarded as “inhibitory ions” in process discussions, which can obscure the fact that they may suppress struvite crystallization through qualitatively different pathways—bulk-phase competition versus surface-controlled inhibition. Without a unified comparison that connects solution speciation, supersaturation, phase evolution, and nucleation kinetics, it is difficult to design robust, mechanism-specific mitigation strategies. This limitation partly explains why struvite recovery performance can vary significantly across wastewater sources even when similar operational parameters are applied.

To address these knowledge gaps, the present study systematically investigates heterogeneous struvite crystallization in the presence of Ca²⁺ and Fe²⁺ across controlled concentration gradients representative of ion-rich wastewater conditions. The concentration ranges examined in this study (Ca²⁺: 0–500 mg/L; Fe²⁺: 0–500 mg/L) encompass levels commonly encountered in anaerobic digestate treatment systems (Ca²⁺: 19–321 mg/L in sludge liquors) and livestock wastewater matrices, where inorganic scaling by calcium-, magnesium-, and phosphate-related compounds has been identified as a critical operational challenge. Batch precipitation experiments are combined with multi-technique characterization (XRD, FTIR, SEM, and XPS) to examine phase composition, bonding environments, and surface chemistry, while real-time monitoring of pH and redox potential provides insight into dynamic solution processes and ensures stable anaerobic conditions for Fe²⁺. Thermodynamic modeling is used to quantify speciation and saturation indices of competing phases, and a kinetic interpretation grounded in classical nucleation theory is adopted to relate observed behaviors to changes in driving force and interfacial parameters. By distinguishing bulk-phase competition from surface-mediated inhibition, this work aims to provide a comparative mechanistic framework to support stable, high-purity struvite recovery from complex wastewaters.

2. Materials and Methods

2.1. Experimental Materials

All chemical reagents used in this study were of analytical grade to ensure consistency and reliability of the experimental results. Reaction solution A (0.25 mol/L Mg²⁺) was prepared by dissolving 61.5018 g of MgSO₄·7H₂O (molecular weight 246.47 g/mol, purity ≥ 99%) in ultrapure water (resistivity ≥ 18.2 MΩ·cm) and diluting to a final volume of 1 L. Reaction solution B (containing 0.25 mol/L PO₄³⁻ and 0.75 mol/L NH₄⁺) was prepared by dissolving 39.0029 g of NaH₂PO₄·2H₂O (molecular weight 156.01 g/mol, purity ≥ 99%) and 40.1254 g of NH₄Cl (molecular weight 53.49 g/mol, purity ≥ 99.5%) in oxygen-depleted water to 1 L. The NH₄⁺/PO₄³⁻ molar ratio of 3:1 was selected based on literature recommendations to maintain sufficient supersaturation for struvite formation [20,21].

Solutions of varying Ca²⁺ concentrations (0–500 mg/L) were prepared using anhydrous CaSO₄ (purity ≥ 98%). Fe²⁺ stock solutions (0–500 mg/L) were prepared using FeSO₄·7H₂O (purity ≥ 99%) immediately before use to minimize oxidation. To prevent Fe²⁺ oxidation, oxygen-depleted water was used throughout the experiment, prepared by purging ultrapure water with high-purity nitrogen (≥99.999%) for at least 30 min. Fe²⁺ solutions were freshly prepared immediately before use, and a positive-pressure nitrogen gas curtain was maintained above the liquid surface throughout the reaction process to ensure anaerobic conditions. All prepared solutions were stored at 4 °C in sealed containers and used within 24 h.

2.2. Experimental Procedure

Batch experiments were conducted to investigate struvite precipitation under controlled conditions. A total of 40 mL of solution A and 40 mL of solution B were transferred into a 1 L volumetric flask. Regulatory factors including Ca²⁺ (0–500 mg/L) and Fe²⁺ (0–500 mg/L) were added to evaluate their effects on precipitation behavior. The mixture was then diluted to 1 L with ultrapure water and transferred to a wide-mouth 1 L Erlenmeyer flask to ensure adequate mixing.

The flasks were sealed with parafilm to limit ammonia volatilization and placed in a constant-temperature orbital shaker (THZ-82 model; Changzhou Guohua Electric Appliance Co., Ltd., Changzhou, China) at 100 rpm for 72 h at 25 ± 1 °C. The shaking conditions were based on preliminary trials to achieve equilibrium while simulating gentle agitation in wastewater environments. The initial pH of the mixed solution was adjusted using 1 mol/L NaOH or HCl. The system reaction then proceeded freely without further pH adjustment to incorporate the proton release kinetics associated with the metabolic and crystallization processes.

After reaction, precipitates were collected by vacuum filtration through a 0.45 µm microporous membrane (Whatman, GE Healthcare, Beijing, China). The solids were rinsed three times with ultrapure water to remove residual ions and loosely attached impurities, then dried in an oven at 40 °C for 48 h to constant weight. Dried samples were stored in a desiccator prior to analysis. All experiments were conducted in triplicate, with control groups excluding Ca²⁺ and Fe²⁺ for baseline comparison. The experimental workflow is illustrated in Figure 1.

2.3. Analytical Methods

A multidimensional characterization protocol was employed to examine the microstructure, surface chemistry, and phase composition of struvite samples under Ca²⁺ and Fe²⁺ interference [22,23]. Dried solid samples were used for scanning electron microscopy (SEM), while portions were finely ground for FTIR, XRD, and XPS analyses to investigate structural features and ion-related effects [24,25].

2.3.1. Fourier Transform Infrared Spectroscopy (FTIR)

Functional group analysis was performed using an Agilent Technologies Cary 630 FTIR spectrometer over the spectral range of 500–4000 cm⁻¹. Samples were analyzed using the attenuated total reflectance (ATR) mode to characterize phosphate-related vibrational features and possible changes associated with Ca²⁺ and Fe²⁺ [26,27]. Characteristic absorption bands including phosphate stretching (1000–1070 cm⁻¹), phosphate bending (560–600 cm⁻¹), NH₄⁺ bending (1400–1500 cm⁻¹), and O–H/N–H stretching (3000–3600 cm⁻¹) were identified and compared across different ion concentrations [28,29].

2.3.2. X-Ray Diffraction (XRD)

Phase identification and crystallographic analysis were performed using a Thermo Fisher Scientific ARL EQUINOX 100 X-ray diffractometer with Cu Kα radiation (λ = 1.54059 Å). Samples were scanned over a 2θ range of 5–90° at a scan rate of 5°/min [30,31]. The diffraction peaks at 2θ ≈ 15.0°, 15.8°, 20.9°, 30.6°, and 33.3° were indexed to the (101), (002), (111), (211), and (022) planes of struvite (MgNH₄PO₄·6H₂O), respectively, according to the standard reference pattern PDF#04-009-6297 (orthorhombic crystal system, space group Pmn2₁) [32,33]. Gaussian fitting was performed on the main diffraction peaks to extract peak positions and full width at half maximum (FWHM). Crystallite sizes were calculated using the Scherrer equation [34]:

$$D={\displaystyle \frac{K\lambda }{\beta \mathit{cos} \theta }}$$

(1)

where D is the crystallite size (nm), K is the Scherrer constant (0.89), λ is the X-ray wavelength (0.154059 nm), β is the FWHM (radians), and θ is the Bragg angle.

2.3.3. X-Ray Photoelectron Spectroscopy (XPS)

Surface elemental composition and chemical states were analyzed using a K-Alpha X-ray photoelectron spectrometer (Thermo Fisher Scientific, Shanghai, China) [35]. The binding energy scale was calibrated using the adventitious C 1s peak at 284.8 eV as an internal reference [36]. High-resolution spectra of C 1s, N 1s, Mg 1s, and P 2p were acquired to evaluate surface chemical environments. For the Ca²⁺ system, the N 1s spectrum was deconvoluted into lattice-bound NH₄⁺ (~401.7 eV) and surface-adsorbed NH₄⁺ (~400.6 eV) components. For the Fe²⁺ system, the P 2p spectrum was further resolved into P–O (134.17 eV) and P=O (133.35 eV) components [37] to investigate Fe–phosphate coordination [38,39].

2.3.4. Scanning Electron Microscopy (SEM)

Surface habit and crystal size were examined using a desktop scanning electron microscope (COXEM EM-30, Daejeon, Republic of Korea). Prior to observation, dried samples were mounted on aluminum stubs using conductive carbon tape and sputter-coated with a thin gold layer to enhance conductivity. Secondary electron imaging was employed to capture surface topography and observe the morphological evolution of struvite crystals under different Ca²⁺ and Fe²⁺ concentrations [40,41].

2.3.5. Real-Time pH Monitoring

Real-time pH monitoring was performed throughout the precipitation process using a calibrated pH electrode (Mettler Toledo InLab Expert Pro, resolution: 0.01 pH units, Beijing Kaishengyuan Technology Co. Ltd., Beijing, China) [42]. The electrode was pre-calibrated with pH 4.00, 7.00, and 10.00 standard buffer solutions (three-point calibration) before each experiment. pH values were recorded at 30 s intervals using an automated data logger (Mettler Toledo SevenExcellence, Greifensee, Switzerland). The electrode probe was positioned at mid-solution depth with constant stirring at 200 rpm to ensure solution homogeneity [43]. Stirring was applied to simulate the mild mixing conditions in practical wastewater treatment systems.

2.3.6. Redox Potential (Eh) Monitoring

The oxidation-reduction potential (Eh) was monitored using a platinum ORP electrode (Mettler Toledo InLab Redox, Ag/AgCl reference). Measured values were converted to the standard hydrogen electrode (SHE) scale by adding 207 mV [44].

The Fe²⁺/Fe³⁺ distribution was calculated using the Nernst equation [45]:

$$Eh=E°+{\displaystyle \frac{RT}{nF}}\times \mathit{ln}{\displaystyle \frac{{Fe}^{3+}}{{Fe}^{2+}}}$$

(2)

where Eh denotes the redox potential of a solution, reflecting the system’s oxidation or reduction capacity; E° = 0.771 V (standard reduction potential for Fe³⁺/Fe²⁺ couple at pH < 6). R is the gas constant; T is the thermodynamic temperature; F is Faraday’s constant; n is the number of electrons transferred in the reaction.

2.3.7. Kinetic Analysis Based on Classical Nucleation Theory (CNT)

To provide mechanistic insights into how Ca²⁺ and Fe²⁺ alter the nucleation landscape of struvite, Classical Nucleation Theory (CNT) was applied to analyze the precipitation kinetics [46,47]. According to CNT, the nucleation rate and critical nucleus size are governed by the supersaturation ratio (S) and interfacial tension (σ). The critical nucleus radius (r*) and Gibbs free energy barrier (ΔG*) were calculated using the following equations [48]:

$$r^{\ast }={\displaystyle \frac{2\sigma v}{kT \mathit{ln}S}}$$

(3)

$$∆G^{\ast }={\displaystyle \frac{16\pi {\sigma }^3{v}^2}{3{\left(kT \mathit{ln}S\right)}^2}}$$

(4)

where v is the molecular volume of struvite (7.89 × 10⁻²⁹ m³, calculated from crystallographic data), $\sigma$ is the interfacial surface tension, k is the Boltzmann constant (1.38 × 10⁻²³ J/K), and T is the absolute temperature (298 K). The supersaturation ratio (S) was calculated from thermodynamic modeling results obtained using Visual MINTEQ 3.1:

$$S=IAP/K_{sp}$$

(5)

where IAP is the ion activity product and K_sp is the solubility product constant of struvite (10^−13.26 at 25 °C). The effective supersaturation under Ca²⁺ interference was estimated by accounting for phosphate consumption by Ca-P co-precipitation, based on the measured decrease in struvite saturation index. The interfacial tension (σ) for the control condition was adopted from literature values for struvite-solution interfaces (σ₀ ≈ 60 mJ/m²). For the Fe²⁺ system, the elevated interfacial tension was estimated based on the observed increase in nucleation barrier, which correlates with surface adsorption effects. The interfacial tension modification was calculated by fitting the experimental crystallite size reduction to the CNT framework, assuming that the primary effect of Fe²⁺ is to increase the effective σ through surface poisoning. It should be noted that the CNT parameters presented here are intended for comparative mechanistic interpretation rather than absolute quantification. The key assumptions include: (1) classical nucleation behavior applies to struvite formation under the studied conditions; (2) the interfacial tension can be approximated as concentration-independent for Ca²⁺ interference but concentration-dependent for Fe²⁺ due to surface adsorption; and (3) the molecular volume remains constant across all conditions. These simplifications are justified for the purpose of distinguishing between thermodynamic (supersaturation-limited) and kinetic (interface-controlled) inhibition mechanisms.

2.3.8. Thermodynamic Modeling

The saturation index (SI) of struvite and potential competing phases was calculated using Visual MINTEQ 3.1 software to provide thermodynamic insights into the precipitation behavior under Ca²⁺ and Fe²⁺ interference [49]. The SI is defined as:

$$SI=\mathit{log}\left({\displaystyle \frac{IAP}{K_{sp}}}\right)$$

(6)

where IAP is the ion activity product and Ksp is the solubility product constant. A positive SI indicates supersaturation and thermodynamic favorability for precipitation, while a negative SI indicates undersaturation. The software accounts for ionic strength effects using the Davies equation [50]:

$$\mathit{log}{\gamma }_i=-A{z}_i^2\left({\displaystyle \frac{\sqrt{I}}{1+\sqrt{I}}}-0.3I\right)$$

(7)

where${ \gamma }_i$ is the activity coefficient, A is the Debye-Hückel constant (0.509 at 25 °C), zᵢ is the ionic charge, and I is the ionic strength.

Input parameters included measured concentrations of Mg²⁺ (10 mM), NH₄⁺ (30 mM), PO₄³⁻ (10 mM), and varying concentrations of Ca²⁺ or Fe²⁺ (0–500 mg/L), along with solution pH (8.5). Temperature was set at 25 °C for all calculations [51]. To ensure valid comparisons across different additive conditions, the initial concentrations of Mg²⁺, NH₄⁺, and PO₄³⁻ were adjusted using Visual MINTEQ 3.1 to maintain the struvite saturation index at SI = 1.85 ± 0.10 in the control condition [52].

3. Results

3.1. Precipitation Performance of Struvite Under Ca²⁺ and Fe²⁺ Interference

Figure 2 presented the variation in total precipitate mass under different concentrations of CaSO₄ and FeSO₄. In the control condition (0 g/L), the precipitate mass was approximately 0.46–0.48 g. As the concentration of CaSO₄ or FeSO₄ increased from 0 to 0.5 g/L, both interference ions caused a gradual increase in the total precipitate mass, but with distinct differences in the magnitude and pattern of change.

For CaSO₄, the total precipitate mass increased slightly in the range of 0–0.3 g/L, rising from approximately 0.48 g to 0.54–0.56 g. A more pronounced increase occurred at higher concentrations: at 0.4 g/L, the mass reached about 0.56 g, and at 0.5 g/L it rose sharply to approximately 0.70–0.72 g, indicating a significant enhancement in precipitation at elevated Ca²⁺ levels. This upward shift suggests that higher CaSO₄ dosages contributed markedly to the increase in solid-phase formation.

For FeSO₄, the overall trend was also an increase in total precipitate mass, but the magnitude of change was smaller than that observed for CaSO₄. From 0 to 0.4 g/L, the mass rose from around 0.46 g to 0.52–0.54 g. At 0.5 g/L, the increase plateaued, maintaining a value close to 0.53 g without a further rise. Across the entire concentration range, the FeSO₄ curve remained consistently lower than the CaSO₄ curve.

A comparison of the two ions demonstrates that Ca²⁺ produced a more substantial enhancement in precipitate formation, especially above 0.3 g/L, with the mass at 0.5 g/L exceeding that of Fe²⁺ by more than 0.15 g. In contrast, Fe²⁺ led to a steadier and more moderate increase, with no abrupt changes at higher concentrations.

Overall, Ca²⁺ and Fe²⁺ both promoted increased precipitate mass within the tested range, but Ca²⁺ exhibited a much stronger effect, particularly at higher dosages. Fe²⁺ showed a more gradual and limited influence. These observations suggest that Ca²⁺ has a greater impact on enhancing solid-phase formation during struvite precipitation than Fe²⁺ under equivalent mass concentrations. General mass gain refers to the total increase in precipitate mass, which includes not only struvite but also other solid phases such as calcium phosphate (CaP) precipitates formed when Ca²⁺ redirects phosphate away from struvite formation.

3.2. Multidimensional Characterization of Struvite Precipitates

3.2.1. FTIR Spectra of Struvite Precipitates Under Ca²⁺ and Fe²⁺ Interference

Figure 3 displayed the FTIR spectra of struvite precipitates obtained under varying concentrations of CaSO₄ and FeSO₄, with HAP serving as a reference standard for comparison. The HAP reference spectrum exhibits characteristic absorption peaks: a broad O–H stretching vibration peak around 3400 cm⁻¹, an asymmetric phosphate stretching vibration peak around 1043 cm₄, and P-O bending vibration peaks in the 560–630 cm₄ region. These peak positions served as diagnostic markers for identifying the formation of Ca–P phases in experimental samples.

At a concentration of 0 g/L, the reference struvite spectrum exhibits characteristic absorption lines of MgNH₄PO₄·6H₂O, including: a strong phosphate stretching vibration band near 1060–1085 cm⁻¹, a bending vibration band around 550–580 cm⁻¹, an NH₄⁺ bending vibration band at approximately 1430–1480 cm⁻¹, and a broad O–H/N–H stretching region spanning 3200–3450 cm⁻¹. These features provided a baseline for evaluating changes induced by Ca²⁺ or Fe²⁺. Several additional absorption features were observed: a band near 1650–1670 cm⁻¹ attributed to the H–O–H bending vibration of crystalline water molecules and the asymmetric bending mode of NH₄⁺; and a weak absorption near 2300–2400 cm⁻¹ corresponding to the combined band or second harmonic vibration of NH₄⁺, partially affected by atmospheric CO₂ interference.

As the CaSO₄ dosage increased, the Fourier Transform Infrared (FTIR) spectrum exhibits progressive changes, indicating the formation of the Ca–P phase. At lower concentrations (0.1–0.2 g/L), the spectrum remains similar to the reference struvite pattern, with phosphate stretching vibration peaks positioned at 1071–1085 cm⁻¹, though peak intensity shows only slight enhancement. However, at higher calcium sulphate concentrations (0.3–0.5 g/L), a marked spectral shift occurs: the phosphate stretching vibration peak progressively shifts from 1060 cm⁻¹ towards 1043 cm⁻¹, approaching the characteristic position of hydroxyapatite (HAP). Concurrently, new absorption features emerge in the 560–630 cm⁻¹ region, closely matching the P–O bond bending vibration peak of the HAP reference material. At a calcium sulphate concentration of 0.5 g/L, the spectrum exhibited pronounced HAP characteristics: the central phosphate band peak shifted to 1043.1 cm⁻¹, with enhanced absorption at 573.8 cm⁻¹ and 621.1 cm⁻¹, showing significant spectral convergence with the HAP standard. The phase composition at this concentration was identified as a mixture of MgNH₄PO₄·6H₂O and Ca₃(PO₄)₂(OH), confirming that elevated Ca²⁺ concentrations promoted bulk Ca–P precipitation through competitive depletion of phosphate.

In contrast, the FeSO₄ series exhibited a markedly different trend. Across all Fe²⁺ concentrations (0.1–0.4 g/L), the spectra retained the characteristic struvite framework structure, with phosphate stretching vibration peaks consistently positioned within the 1066–1081 cm⁻¹ range-a distinct deviation from the HAP reference peak at 1043 cm⁻¹. Peak intensities increased with rising FeSO₄ dosage, most notably within the O–H/N–H stretching region (3389–3424 cm⁻¹), though no convergence towards the HAP spectral pattern was observed. Unlike the CaSO₄ system, no novel absorption features associated with crystalline Fe–P phases emerged at low to moderate Fe²⁺ concentrations. Only at an FeSO₄ concentration of 0.5 g/L did the phosphate band exhibit a slight shift towards 1043 cm⁻¹, accompanied by a faint characteristic peak in the 570–620 cm⁻¹ region. This suggests the potential formation of amorphous or surface-associated Fe–P species at high concentrations, though without the distinct crystalline HAP-type features observed in the CaSO₄ system.

Overall, FTIR results interpreted against the HAP standard spectrum revealed distinctly different interference mechanisms for Ca²⁺ and Fe²⁺. The CaSO₄ system exhibits concentration-dependent spectral evolution towards HAP characteristics: beyond 0.3 g/L, the phosphate band shifts markedly and diagnostic P–O bending vibration peaks emerge, providing direct spectroscopic evidence for bulk Ca–P precipitation. Conversely, the FeSO₄ system consistently retained the struvite spectral framework at most concentrations without approaching the HAP reference pattern. This aligns with the surface-mediated inhibition mechanism: Fe²⁺ adsorbs onto the struvite surface to form an amorphous Fe–P layer, rather than crystalline bulk precipitates.

3.2.2. XRD Patterns of Struvite Precipitates Under Ca²⁺ and Fe²⁺ Interference

FTIR was used to characterize struvite precipitates formed in the presence of CaSO₄. The results show that at lower concentrations (0.1–0.2 g/L), the characteristic absorption features of struvite were largely preserved. However, at higher concentrations (≥0.3 g/L), progressive spectral shifts toward HAP characteristics were observed, indicating Ca-P phase formation. Figure 4 and Figure 5 presents the XRD patterns of heterogeneously nucleated struvite. The diffraction peaks located at 2θ values of approximately 15.0°, 15.8°, 20.9°, 30.6°, and 33.3° can be indexed to the (101), (002), (111), (211), and (022) planes of struvite (MgNH₄PO₄·6H₂O), respectively, according to the standard reference file PDF#04-009-6297. The identified phase belongs to the orthorhombic crystal system with the space group Pmn2₁ (No. 31).

Table 1. Ca²⁺and Fe²⁺ system—Main peak2θ position (°).

Crystal Face	Control	0.1 g/L		0.2 g/L		0.3 g/L		0.4 g/L		0.5 g/L
		Ca	Fe	Ca	Fe	Ca	Fe	Ca	Fe	Ca	Fe
(101)	15.00	15.06	14.86	15.08	14.88	15.06	14.92	14.98	14.90	14.98	14.92
(002)	15.80	15.86	15.68	15.86	15.68	15.86	15.72	15.80	15.68	15.78	15.70
(111)	20.86	20.92	20.74	20.90	20.74	20.92	20.74	20.84	20.74	20.86	20.76
(211)	30.62	30.68	30.50	30.68	30.52	30.68	30.50	30.58	30.52	30.60	30.52
(022)	33.26	33.30	33.14	33.32	33.16	33.30	33.14	33.26	33.16	33.24	33.16

presents the main peak positions in 2θ degrees for the Ca²⁺ and Fe²⁺ system during phosphorus recovery from simulated wastewater struvite crystallization.

Within the Ca²⁺ concentration gradient of 0–0.5 g/L, no significant changes in the phase structure of the samples are observed. Moreover, under Ca²⁺ concentrations ranging from 0.1 to 0.5 g/L, the positional shift of the main diffraction peak of struvite at 20.9° corresponding to the (111) plane is very small (Δ2θ < 0.09°), and no noticeable variation in lattice parameters is detected, indicating that Ca²⁺ ions are not incorporated into the struvite crystal lattice.

By calculating the full width at half maximum (FWHM) of the main diffraction peak, the results show that, with increasing Ca²⁺ concentration from 0 to 0.5 g/L, the (111) diffraction peak gradually broadens, with the FWHM increasing from 0.191° to 0.238°. This broadening is likely attributed to both a reduction in crystallite size and a decrease in crystallinity degree, as confirmed by the progressive decrease in diffraction peak intensity and increase in peak asymmetry with increasing Ca²⁺ concentration. The combined effect of smaller crystallite size and lower crystallinity indicates that Ca²⁺ competition for phosphate disrupts the ordered arrangement of struvite crystal lattices, leading to more disordered crystal structures. Crystallite sizes estimated using the Scherrer equation further confirm this trend: as the Ca²⁺ concentration increases from 0 to 0.5 g/L, the average crystallite size of struvite decreases from 63.0 nm to 48.7 nm.

These results indicated that, during struvite precipitation, the presence of Ca²⁺ ions competes with Mg²⁺ ions for phosphate (PO₄³⁻), particularly at higher Ca²⁺ concentrations, thereby significantly inhibiting struvite formation. This competitive interaction affects the size, habit, leading to smaller crystallites. The reaction predominantly involves the combination of Ca²⁺ with PO₄³⁻ to form calcium phosphate phases rather than the incorporation of Ca²⁺ into the struvite lattice.

In-depth X-ray diffraction (XRD) analysis reveals that the diffraction peaks located at 2θ values of approximately 15.0°, 15.8°, 20.9°, 30.6°, and 33.3° can be indexed to the (101), (002), (111), (211) and (022) planes of struvite (MgNH₄PO₄·6H₂O), respectively. According to the standard reference file PDF#04-009-6297, the phase belongs to the orthorhombic crystal system with the space group Pmn2₁ (No. 31). The introduction of trace amounts of Fe²⁺ does not significantly alter the phase structure of the samples.

Within the FeSO₄ concentration gradient of 0–0.5 g/L, the positional shift of the main diffraction peak of struvite at 20.9° corresponding to the (111) plane is very small (Δ2θ < 0.15°), and no lattice distortion is observed, indicating that Fe²⁺ ions are not incorporated into the struvite crystal lattice. However, analysis of the full width at half maximum (FWHM) and the corresponding crystallite size reveals that, with increasing Fe²⁺ concentration, the main diffraction peak gradually broadens, with the FWHM increasing from 0.191° to 0.239°, while the crystallite size of struvite decreases from 63.0 nm to 49.8 nm. These results indicate that the presence of Fe²⁺ reduces the size of the crystallite.

The interaction between Fe²⁺ and the struvite surface does not significantly modify the Crystal habit and therefore does not markedly affect the crystal growth mode. Fe²⁺ may compete with Mg²⁺ for PO₄³⁻, potentially affecting struvite formation. As the Fe²⁺ concentration increases, Fe²⁺ increasingly binds with PO₄³⁻ to generate ferrous phosphate, leading to the preferential consumption of phosphate by Fe²⁺ and a consequent reduction in phosphorus recovery efficiency. This trend becomes more pronounced with increasing Fe²⁺ concentration. However, because the Fe²⁺ concentration remains extremely low, the amount of ferrous phosphate formed is insufficient to be detected by XRD and is therefore neglected in the diffraction analysis.

3.2.3. SEM Observation of the Morphological Evolution of Struvite Crystals

Figure 6 showed the SEM images of heterogeneous nucleated struvite crystals obtained under different CaSO₄ and FeSO₄ concentrations. In the control group (0 g/L), the products exhibited the characteristic orthorhombic prismatic habit of struvite, with well-defined edges, smooth facets, and relatively isolated crystals. This typical habit is consistent with previously reported SEM observations of pure struvite precipitates.

With increasing CaSO₄ concentration, the habit of the precipitates remained recognizably struvite across the entire range but showed gradual deviations from the well-formed prisms observed in the control. At 100–200 mg/L, prismatic struvite crystals were still abundant, although fragmentation increased and the particle size distribution became more heterogeneous. At 300 mg/L, the samples contained both intact crystals and a noticeable number of broken or irregular pieces, leading to a more complex surface appearance. At 400–500 mg/L, increased aggregation and fragmentation were observed, with a higher proportion of plate-like debris mixed with prismatic struvite. Similar trends in morphological broadening and increased irregularity under Ca²⁺ conditions have been reported in earlier studies examining the influence of Ca²⁺ on struvite crystallization [10,12].

The FeSO₄ series displayed a comparable pattern of progressive heterogeneity. At 100 mg/L, the characteristic prismatic habit was still clearly visible, although small fragments and irregular pieces appeared more frequently than in the control. At 200–300 mg/L, the surfaces became increasingly populated with fractured crystals and irregularly shaped particles, reflecting a general reduction in crystal uniformity. At 400–500 mg/L, many prismatic crystals persisted, but the overall habit was dominated by heterogeneous aggregates and numerous small, broken pieces. The increasing fragmentation and loss of uniformity under Fe-containing conditions are consistent with previously documented effects of metal ions on struvite crystal habit and aggregation behavior [11,28].

Across both ion series, the SEM images indicate that the characteristic struvite habit remained present throughout the tested concentration range, but higher concentrations of Ca²⁺ or Fe²⁺ resulted in more pronounced fragmentation, increased irregularity, and a broader distribution of particle habit and sizes. These changes reflect a general disturbance of crystal development rather than the formation of distinctly new morphologies.

EDS point analysis was performed on the samples to quantify the elemental composition (Figure 6 and Table S1). In the CaSO₄ system, the Ca/Mg mass ratio was approximately 3.5 under 500 mg/L Ca²⁺ conditions. Although no crystalline Ca-P phase was detected by XRD, thermodynamic calculations indicated that ACP was significantly supersaturated (SI = 2.05); thus, the high Ca content was attributed to the amorphous calcium phosphate (ACP) phase co-precipitated with struvite. In the FeSO₄ system, the Fe/Mg mass ratio was approximately 4 under 500 mg/L Fe²⁺ conditions, and the Fe content in the high-brightness areas of the SEM images exceeded 25 wt%, indicating a non-uniform Fe distribution. This is consistent with the surface adsorption mechanism: Fe²⁺ preferentially adsorbs onto the struvite surface to form a localized amorphous Fe-P deposition layer, a phenomenon further confirmed by the P–O–Fe coordination signal detected by XPS.

SEM-EDS semi-quantitative analysis of solid-phase products provides supporting evidence for Fe²⁺ inhibiting struvite formation. As FeSO₄ concentration increased from 0 to 500 mg/L, Mg content in the precipitate decreased from approximately 21.15 wt% to 6.60 wt%, with a corresponding increase in Fe content. This compositional shift indicates that Fe²⁺ interference reduced the proportion of struvite within the solid phase, with a portion of phosphate being fixed as amorphous Fe-P deposits. It should be noted, however, that EDS analysis is a semi-quantitative method reflecting elemental distribution only in localised regions. Future studies may employ ICP-OES or ICP-MS for precise quantification of acid-digested solid-phase products, establishing a complete mass balance to further validate the proposed inhibition mechanism.

3.2.4. XPS Characterization of Surface Element Composition and Chemical State

In calcium sulfate solutions with a concentration gradient of 0–0.5 g/L, the X-ray photoelectron spectroscopy (XPS) survey spectra of heterogeneously nucleated struvite consistently exhibit characteristic peaks of N 1s (401.5 eV), P 2p (133.3 eV), and Mg 1s (1304.24 eV). It should be noted that graphite was used to calibrate the instrument during XPS measurements, which inevitably introduced carbon contamination. Accordingly, three sets of peaks were observed in the high-resolution C 1s spectra of all samples (Figure 7(Aa)), corresponding to sp²-hybridized carbon (C–C) in the graphite structure, C–O bonds, and C=O bonds, respectively [53]. To facilitate comparison of the effects of different calcium sulfate concentrations on the heterogeneous nucleation of struvite, the C–C bond at 284.8 eV was used as an internal reference to calibrate all subsequent elemental spectra [54].

As shown in Figure 7(Ab), struvite synthesized in the absence of CaSO₄ exhibits two N 1s peaks at 401.73 and 400.65 eV, which can be assigned to NH₄⁺ ions incorporated in the struvite lattice and free NH₄⁺ species adsorbed on the surface, respectively [55]. With increasing calcium sulfate concentration, the peak area corresponding to lattice-bound NH₄⁺ gradually decreases, whereas the peak area associated with free NH₄⁺ increases slightly. This phenomenon results from reduced struvite crystallinity: Ca²⁺ competition for PO₄³⁻ lowers the crystallization driving force, producing less ordered lattices that incorporate NH₄⁺ less effectively, thereby increasing the proportion of surface-adsorbed NH₄⁺. The relatively small magnitude of this variation indicates that Ca²⁺ ions mainly influence struvite crystallization through competitive interactions in the solution phase rather than through direct interactions with the crystal surface [56].

Figure 7(Ac) illustrates the evolution of the Mg 1s spectrum for heterophase-nucleated struvite as a function of calcium sulphate concentration. In the absence of CaSO₄, the Mg 1s spectrum of struvite resolves into two distinct peaks at 1305.13 eV and 1304.24 eV. The peak at 1305.13 eV is attributed to the hydrated magnesium ion Mg(H₂O)₆²⁺ forming on the struvite surface [57]. while the peak at 1304.24 eV corresponds to the Mg–O bonds within the struvite crystal structure. This double-peak configuration reflects the coexistence of surface-hydrated magnesium ions and lattice-bound magnesium within the pristine struvite sample. As the calcium sulphate concentration increases, the relative peak area corresponding to the hydrated magnesium ions (Mg(H₂O)₆²⁺) at ~1305 eV progressively increases, while the peak area corresponding to the structural Mg–O bonds correspondingly decreases. This systematic shift in the Mg 1s spectrum profile indicates that Ca²⁺ ions progressively alter the chemical environment of Mg²⁺ during heterogeneous nucleation [58,59]. The increased proportion of surface-hydrated magnesium suggests that Ca²⁺ competition with phosphate reduces the effective binding of Mg²⁺ into the struvite lattice. Consequently, a higher proportion of Mg²⁺ remains in a surface-hydrated state rather than integrating into the crystal structure.

The high-resolution P 2p spectra (Figure 7(Ad)) exhibit a single broad peak corresponding to PO₄³⁻ groups within the struvite structure. With increasing calcium sulfate concentration, the binding energy of the phosphate peak remains stable at approximately 133.54 eV, suggesting that phosphate groups reside in a similar chemical environment across all samples. This stability can be attributed to the limited influence of Ca²⁺ ions on the surface electron density of phosphorus atoms during the heterogeneous nucleation of struvite [60].

We further investigated the influence of ferrous sulfate solutions with a concentration gradient of 0–0.5 g/L on the heterogeneous nucleation of struvite by X-ray photoelectron spectroscopy (XPS). The C–C bond at 284.8 eV in the high-resolution C 1s spectra (Figure 7(Ba)) was again used as an internal reference to calibrate all subsequent elemental spectra.

Struvite synthesized in the absence of FeSO₄ exhibits an N 1s high-resolution XPS spectrum (Figure 7(Bb)) similar to that of samples prepared in CaSO₄ solutions. The peaks at 401.67 and 399.98 eV are assigned to NH₄⁺ incorporated in the struvite lattice and free NH₄⁺ species adsorbed on the surface, respectively. With increasing ferrous sulfate concentration, the relative content of lattice-bound NH₄⁺ gradually decreases, while the relative content of free NH₄⁺ correspondingly increases. This phenomenon is attributed to the adsorption of Fe²⁺ ions at the crystal–solution interface, which interferes with the normal crystallization process and hinders the effective incorporation of NH₄⁺ into the newly formed lattice [61]. The high-resolution XPS spectrum of Mg 1s in the blank system (without FeSO₄) (Figure 7(Bc)) exhibits spectral characteristics similar to those of the CaSO₄ system. This spectrum can also be resolved into two distinct peaks: the peak at ~1305 eV corresponds to hydrated magnesium ions (Mg(H₂O)₆²⁺) on the struvite surface, while the peak at ~1304 eV is attributed to the Mg–O bonds within the struvite lattice structure. With increasing ferrous sulphate concentration, the relative peak area of the hydrated magnesium component gradually increases, while that of the lattice-bound Mg–O component progressively decreases. This trend indicates that Fe²⁺ adsorption at the crystal-solution interface interferes with the normal incorporation of Mg²⁺ into the struvite lattice, resulting in a larger proportion of magnesium remaining in a surface-hydrated state.

To further elucidate the effect of ferrous sulfate on the heterogeneous nucleation of struvite, the P 2p spectra corresponding to PO₄³⁻ were deconvoluted into P–O (134.17 eV) and P=O (133.35 eV) components (Figure 7(Bd)) [62]. As the ferrous sulfate concentration increases from 0.2 to 0.5 g/L, significant variations in both the peak positions and peak areas of the P–O and P=O components are observed among different samples. Previous studies have shown that the heavy metal ion Fe²⁺ can coordinate with the P–O bonds in PO₄³⁻, and with increasing Fe²⁺ concentration, the relative peak area corresponding to the P–O bond gradually increases, indicating an increase in the relative Fe content [63]. The coordination of Fe²⁺ with P–O bonds to form P–O–Fe linkages [64] induces a neighboring-atom effect that alters the surface electron density of phosphorus atoms, thereby causing the binding energies of the P–O and P=O components to shift progressively toward lower values [65,66].

3.3. Real-Time Monitoring of Solution Chemistry

3.3.1. pH Evolution During Crystallization

Figure 8 presented the pH-time profiles during struvite crystallization under different interference conditions. In the control group (0 g/L additive), pH exhibited a moderate decrease from 9.20 to 8.65 (ΔpH = 0.55) over 120 min, characteristic of stoichiometric struvite crystallization where proton release accompanies the incorporation of NH₄⁺ and PO₄³⁻ into the crystal lattice.

The addition of Ca²⁺ induced more pronounced acidification. At 0.3 g/L CaSO₄, the pH decreased from 9.20 to 8.40 (ΔpH = 0.80), representing a 45% increase in proton release compared to the control. This enhanced acidification was progressive and sustained throughout the crystallization period, consistent with continuous phosphate consumption by competing Ca-P precipitation reactions occurring simultaneously with struvite formation.

In contrast, the Fe²⁺ system exhibited distinctly different pH kinetics. At 0.3 g/L FeSO₄, the initial pH decline was more rapid and severe, with the pH dropping from 9.20 to 8.25 (ΔpH = 0.95) within the first 30 min. This initial steep decline followed by a gradual levelling off indicates a rapid surface-mediated reaction rather than bulk phase precipitation. This biphasic pH curve is characteristic of adsorptive surface processes: the initial rapid proton release corresponds to saturation of finite crystal surface sites, with subsequent reaction rates constrained by new surface exposure or diffusion within the adsorbed layer. In contrast, bulk precipitation exhibits a continuous, progressive pH decline proportional to the sustained depletion of phosphate. The pronounced early-stage proton release is consistent with the formation of amorphous Fe-P surface layers through in situ interfacial reactions.

3.3.2. Redox Potential (Eh) and Iron Speciation

Table 2. Monitoring of Redox Potential and Iron Speciation.

Time (min)	pH	Eh (mV vs. SHE)	Fe2+ (mM)	Fe3+ (mM)	Fe2+ Fraction (%)
0	9.20	−185 ± 5	5.36	<0.01	>99.9
5	8.80	−175 ± 8	4.65	<0.01	>99.9
15	8.52	−160 ± 6	3.82	<0.01	>99.9
30	8.35	−145 ± 5	2.95	<0.01	>99.9
60	8.18	−130 ± 5	2.15	<0.01	>99.9
120	8.05	−115 ± 5	1.45	<0.01	>99.9

presented the redox potential (Eh) monitoring results and iron speciation analysis throughout the crystallization process in the Fe²⁺ system (0.3 g/L FeSO₄). The Eh values remained consistently within the reducing range (−185 to −115 mV vs. SHE) throughout the 120 min reaction period, confirming that anaerobic conditions were successfully maintained.

The dissolved Fe³⁺ concentration remained below the detection limit (<0.01 mM) throughout the experiment, with Fe²⁺ accounting for >99.9% of total dissolved iron. The progressive decrease in dissolved Fe²⁺ concentration (from 5.36 to 1.45 mM) reflects its consumption through precipitation and/or surface adsorption reactions. These results validate that the observed interference effects can be attributed specifically to Fe²⁺ species, excluding Fe³⁺ oxidation as a confounding factor.

3.4. Thermodynamic Saturation Index Calculations

Thermodynamic modeling using Visual MINTEQ 3.1 was performed to calculate the saturation indices (SI) of struvite and potential competing mineral phases under varying Ca²⁺ and Fe²⁺ concentrations. The results are summarized in

Table 3. Saturation indices of mineral phases calculated by Visual MINTEQ 3.1. under Fe²⁺ interference.

Condition	SI (Struvite)	SI (Vivianite)	SI (Hydroxyapatite)	SI (Fe(OH)3)
Control, pH 8.5	1.85 ± 0.08	-	2.12 ± 0.15	-
0.1 g/L Fe, pH 8.5	1.72 ± 0.09	3.45 ± 0.12	2.08 ± 0.14	5.82 ± 0.20
0.2 g/L Fe, pH 8.5	1.65 ± 0.09	3.85 ± 0.14	2.02 ± 0.15	6.15 ± 0.21
0.3 g/L Fe, pH 8.5	1.58 ± 0.10	4.21 ± 0.15	1.95 ± 0.16	6.45 ± 0.22
0.4 g/L Fe, pH 8.5	1.52 ± 0.11	4.45 ± 0.16	1.88 ± 0.17	6.68 ± 0.24
0.5 g/L Fe, pH 8.5	1.45 ± 0.12	4.68 ± 0.18	1.82 ± 0.18	6.89 ± 0.25

and

Table 4. Saturation indices of mineral phases calculated by Visual MINTEQ 3.1. under Ca²⁺ interference.

Condition	SI (Struvite)	SI (Hydroxyapatite)	SI (ACP)	SI (Brushite)
Control, pH 8.5	1.85 ± 0.08	2.12 ± 0.15	−0.45 ± 0.10	−1.25 ± 0.12
0.1 g/L Ca, pH 8.5	1.68 ± 0.08	3.85 ± 0.20	0.82 ± 0.12	−0.58 ± 0.10
0.2 g/L Ca, pH 8.5	1.60 ± 0.09	4.18 ± 0.21	1.15 ± 0.13	−0.32 ± 0.11
0.3 g/L Ca, pH 8.5	1.52 ± 0.10	4.52 ± 0.22	1.48 ± 0.14	−0.05 ± 0.11
0.4 g/L Ca, pH 8.5	1.45 ± 0.10	4.85 ± 0.24	1.78 ± 0.15	0.18 ± 0.12
0.5 g/L Ca, pH 8.5	1.38 ± 0.11	5.15 ± 0.25	2.05 ± 0.16	0.42 ± 0.13

.

In both systems, the struvite SI decreased progressively with increasing interferent concentration, indicating reduced thermodynamic driving force for struvite precipitation. In the Fe²⁺ system, vivianite (Fe₃(PO₄)₂·8H₂O) exhibited high supersaturation (SI = 3.45–4.68) despite the absence of crystalline Fe-P phases in XRD. In the Ca²⁺ system, hydroxyapatite (HAP) SI increased dramatically (2.12 → 5.15), consistent with the HAP reflections observed in XRD at elevated Ca²⁺ concentrations.

4. Discussion

To elucidate the distinct inhibition pathways observed in this study, the proposed interference mechanisms of Ca²⁺ and Fe²⁺ on struvite crystallization are summarized in Figure 9. As illustrated in the schematic, the interactions can be fundamentally categorized into two distinct modes: bulk-phase processes and surface-controlled processes. The left panel of Figure 9 highlights how Ca²⁺ primarily exerts its influence in the bulk solution through hydration competition and ion pairing, which hinders the nucleation of struvite. In contrast, the right panel depicts Fe²⁺ interference as being dominated by surface-controlled mechanisms, including adsorption onto the crystal facets, surface oxidation, and the subsequent formation of an amorphous coating. This conceptual framework serves as a basis for interpreting the experimental results detailed below.

4.1. Influence of Ca²⁺: Bulk-Phase Competition and Formation of Ca–P Phases

The combined experimental results demonstrate that Ca²⁺ interferes with struvite crystallization primarily through bulk-phase competition for phosphate. As CaSO₄ concentration increased, although the thermodynamic model predicts the existence of HAP, due to low crystallinity or content, XRD failed to detect HAP reflections. However, thermodynamic calculations (Table 4): Visual MINTEQ simulations indicate that as Ca²⁺ concentration increases from 0 to 0.5 g/L, the saturation index for ACP rises from −0.45 to 2.05, while the saturation index for hydroxyapatite (HAP) increased from 2.12 to 5.15, both reaching significantly supersaturated states. This indicates that Ca-P precipitation is thermodynamically favourable. Changes in precipitate mass (Figure 2): Total precipitate mass increased from 0.48 g in the control group to 0.72 g at 0.5 g/L Ca²⁺. The increment of approximately 0.24 g cannot be explained solely by struvite, indicating the formation of additional Ca-P solid phases. EDS analysis (Figure 6): Under 500 mg/L Ca²⁺ conditions, the Ca/Mg mass ratio reached approximately 3.5, confirming substantial Ca incorporation into the solid-phase products. XRD crystallinity decline: The FWHM of the struvite (111) peak increased from 0.191° to 0.238°, while grain size decreased from 63.0 nm to 48.7 nm. This indicates phosphate is competitively consumed by the Ca-P phase, diminishing the crystallisation drive for struvite. The above evidence indicates that the Ca-P phase does indeed form, but its amorphous nature renders it undetectable by XRD. The sustained pH decrease, consistent with the kinetics of proton release during Ca-P coprecipitation, provides indirect confirmation of Ca-P phase formation. Together with the suppression of struvite peaks, especially above 300 mg/L.These observations suggest that Ca²⁺ redirects phosphate toward Ca–P precipitation pathways, consistent with established dissolution–precipitation mechanisms. Thermodynamic calculations indicate that rapid phosphate depletion in the presence of Ca²⁺ correlates with the formation of amorphous calcium phosphate (ACP) intermediates, which act as kinetic traps diverting material flow away from struvite nucleation.

SEM analyses corroborated this trend, showing a transition from well-defined prismatic struvite crystals at low Ca²⁺ concentrations to rounded, aggregated, or amorphous structures as Ca²⁺ increased. The absence of significant Ca incorporation in XPS further supports the interpretation that Ca²⁺ affects struvite formation mainly through solution-phase interactions rather than lattice substitution.

The thermodynamic modeling results provide quantitative support for the bulk-phase competition mechanism. As Ca²⁺ concentration increased from 0 to 0.5 g/L, the struvite SI decreased by 25% (from 1.85 to 1.38), while the HAP SI increased by 143% (from 2.12 to 5.15). The ACP SI transitioned from undersaturation (−0.45) to significant supersaturation (2.05), consistent with the stepwise precipitation pathway: dissolved Ca²⁺ + PO₄³⁻ → ACP → crystalline HAP. The pH monitoring data corroborate this mechanism: the progressive and sustained acidification (ΔpH = 0.80 at 0.3 g/L Ca²⁺) reflects continuous proton release from co-precipitation reactions, without the rapid initial pH drop characteristic of surface-mediated processes.

4.2. Influence of Fe²⁺: Surface Adsorption, Growth Poisoning, and Amorphous Fe–P Layer Formation

Fe²⁺ exhibited a mechanistically distinct pathway characterized by surface-controlled inhibition. Across all FeSO₄ concentrations tested, struvite remained the only crystalline phase detectable by XRD, indicating that Fe–P species were amorphous or poorly crystalline. SEM imaging revealed pronounced surface modification even at low Fe²⁺ concentrations, with feather-like, dendritic, and ultimately gel-like deposits enveloping struvite crystals [34,63].

XPS provided additional evidence for Fe-mediated surface interactions. The substantial positive shift in the P–OH binding energy at moderate Fe²⁺ concentrations suggests strong coordination between iron species and phosphate surface groups. High-resolution structural studies from the literature show that the (001) face of struvite exposes multiple outward-oriented oxygen atoms per phosphate group, rendering it highly susceptible to metal binding and surface poisoning [33,37]. Previous work further reports that Fe³⁺ can rapidly deteriorate struvite habit and inhibit ordered crystal growth more strongly than Ca²⁺, consistent with the severe morphological degradation observed in this study [8,11].

The Fe²⁺ system presents a revealing thermodynamic paradox. Despite extremely high struvite supersaturation (SI = 3.45–4.68), no crystalline Fe-P phases were detected, indicating that kinetic factors—specifically rapid surface adsorption—dominate over thermodynamic precipitation. The pH kinetics provide direct evidence: the steep initial pH decline (ΔpH = 0.70 within 30 min) at 0.3 g/L Fe²⁺ contrasts sharply with the gradual acidification in the Ca²⁺ system, consistent with rapid interfacial reactions. Eh monitoring confirmed that these effects are attributable to Fe²⁺ specifically, with Fe³⁺ remaining below detection limits throughout the experiment. The progressive consumption of dissolved Fe²⁺ (from 5.36 to 1.45 mM) without corresponding crystalline Fe-P phase formation strongly supports the conclusion of surface adsorption and amorphous Fe-P layer formation. This is confirmed by XPS analysis: the P 2p spectrum shows a gradual increase in the relative peak area of P–O bonds (from 30.2% to 45.7% at 0.5 g/L Fe²⁺) and a downward shift in binding energy (from 134.17 eV to 133.23 eV), indicating the formation of P–O–Fe linkages on the struvite surface [58,59].

4.3. Kinetic Analysis Based on Classical Nucleation Theory (CNT)

To provide deeper mechanistic insights into how Ca²⁺ and Fe²⁺ alter the nucleation landscape of struvite, Classical Nucleation Theory (CNT) was applied to analyze the precipitation kinetics. The CNT framework is applied here as a comparative tool to distinguish between thermodynamic and kinetic inhibition mechanisms, rather than to provide absolute quantitative predictions. The supersaturation values were derived from thermodynamic modeling (Visual MINTEQ 3.1), while the interfacial tension for the control was adopted from published literature values. For the Fe²⁺ system, the elevated interfacial tension reflects the cumulative effect of surface adsorption processes and was estimated by fitting experimental observations to the CNT framework. Despite these simplifications, the CNT analysis provides valuable mechanistic insights by quantitatively linking the observed crystallization behavior to changes in either driving force (ln S) or interfacial resistance (σ).

Table 5. Kinetic parameters for struvite nucleation derived from classical nucleation theory (CNT) under ion interference.

Parameter	Symbol	Unit	Control Group (0 g/L)	Ca2+ Functional Group (0.5 g/L)	Fe2+ Functional Group (0.3 g/L)
Supersaturation	S	-	8.54	4.12	8.30
Interfacial tension	σ	mJ/m2	60.00	62.5	85.00
Critical nuclear radius	r*	nm	0.85	1.62	1.48
Gibbs free energy barrier	ΔG	×10−19 J	1.82	8.45	6.88
Critical Growth Unit	n*	-	15.00	45.00	32.00

summarizes the kinetic parameters derived from the Classical Nucleation Theory (CNT) for struvite nucleation under various ion interference conditions.

In the control condition, the high supersaturation (S ≈ 8.54) and moderate interfacial tension (σ ≈ 60 mJ/m²) result in a low energy barrier (ΔG* = 1.82 × 10⁻¹⁹ J) and small critical nucleus (r* ≈ 0.85 nm, n* ≈ 15 growth units). This thermodynamically favorable condition enables facile nucleation and ordered crystal growth.

The addition of Ca²⁺ (0.5 g/L) dramatically reduces effective supersaturation to S ≈ 4.12 through phosphate consumption by Ca-P co-precipitation. According to CNT, this decreased driving force (ln S) increases the Gibbs free energy barrier nearly fourfold (ΔG* = 8.45 × 10⁻¹⁹ J). Consequently, the critical nucleus size expands to approximately 45 growth units (n*), requiring larger local concentration fluctuations to trigger stable nucleation. This ‘thermodynamic starvation’ mechanism favors the formation of poorly crystalline or amorphous phases, supported by multiple lines of evidence: (1) XRD results show increased FWHM of struvite diffraction peaks (from 0.191° to 0.238°) and decreased crystallite size (from 63.0 nm to 48.7 nm) with increasing Ca²⁺ concentration; (2) SEM images reveal distorted, fragmented crystals and aggregated structures at high Ca²⁺ levels; (3) thermodynamic calculations indicate a 25% decrease in struvite supersaturation (SI from 1.85 to 1.38) and significant supersaturation of ACP (SI = 2.05 at 0.5 g/L Ca²⁺). This is consistent with previous studies reporting that reduced supersaturation during crystallization leads to incomplete lattice arrangement and the formation of poorly crystalline phases [26,47]. In contrast, Fe²⁺ exerts minimal influence on bulk solution thermodynamics (S remains high at 8.30), but profoundly alters interfacial characteristics. Strong adsorption of Fe species onto nucleating clusters elevates the interfacial tension to σ ≈ 85 mJ/m². Since ΔG* is proportional to σ³, even moderate increases in interfacial tension cause dramatic increases in the nucleation barrier (ΔG* = 6.88 × 10⁻¹⁹ J). This mechanism-termed “surface poisoning”-disrupts the lattice arrangement of octahedral [Mg(H₂O)₆]²⁺ growth units, preventing their ordered integration into the crystal lattice. The result is the amorphous Fe-P surface layer observed in SEM and XPS analyses.

These CNT calculations reveal fundamentally different inhibition pathways: Ca²⁺ reduces the thermodynamic driving force (S↓) via (lnS)², whereas Fe²⁺ increases ΔG* through the σ³ term, thereby elevating interfacial tension (σ↑). This distinction explains why Ca²⁺ forms bulk Ca-P precipitates, while Fe²⁺ generates an amorphous surface layer without distinct crystalline phases.

4.4. Comparative Mechanistic Framework: Thermodynamic Limitation Versus Kinetic Inhibition

The preceding analyses reveal that Ca²⁺ and Fe²⁺ suppress struvite crystallization through fundamentally distinct pathways (

Table 6. Comparative summary of Ca²⁺ and Fe²⁺ inhibition mechanisms on struvite crystallization.

Characteristic	Ca2+	Fe2+
Primary mechanism	Bulk-phase competition	Surface-controlled inhibition
Mode of action	Phosphate sequestration into Ca-P phases	Adsorption onto crystal surfaces
Effect on supersaturation (S)	Significant reduction (↓ 47%)	Minimal change
Effect on interfacial tension (σ)	Slight increase	Significant increase (↑ 42%)
Secondary phases	Amorphous/crystalline Ca-P	Amorphous Fe-P surface layer
pH response	Gradual, sustained acidification	Rapid initial pH drop
Mitigation strategy	Mg:Ca ratio adjustment; Ca pre-precipitation	Surface-active additives; redox control

). Ca²⁺ acts primarily through bulk-phase thermodynamic competition: by sequestering phosphate into amorphous or crystalline calcium phosphate phases, it reduces solution supersaturation (S↓), thereby diminishing the driving force for struvite nucleation. Within the CNT framework, this mechanism operates through the (ln S)² term, requiring larger concentration fluctuations to achieve critical nucleus formation.

In contrast, Fe²⁺ exerts surface-controlled kinetic inhibition: rather than significantly altering bulk supersaturation, Fe²⁺ adsorbs onto active growth sites and increases the crystal–solution interfacial energy (σ↑). This elevates the nucleation energy barrier through the σ³ term, retarding both nucleation and crystal growth without producing detectable crystalline secondary phases.

This mechanistic distinction—thermodynamic limitation versus kinetic inhibition—has direct implications for struvite recovery process design. Ca²⁺ interference, driven by phosphate competition, may be mitigated by increasing Mg:P ratios, implementing Ca pre-precipitation, or operating at pH conditions that disfavor Ca-P nucleation. Fe²⁺ interference, driven by surface poisoning, may require alternative strategies such as surface-active additives that compete for adsorption sites, strict anaerobic control to prevent Fe³⁺ formation, or Fe²⁺ pre-removal via sulfide precipitation. Recognizing these distinct pathways enables the development of targeted, mechanism-specific mitigation approaches rather than generalized “inhibitory ion” treatments, ultimately supporting more robust and predictable struvite recovery from complex wastewater matrices.

5. Conclusions

This study systematically compared the influence of Ca²⁺ and Fe²⁺ on heterogeneous struvite nucleation and crystallization using complementary characterization, solution monitoring, thermodynamic evaluation, and nucleation-kinetic interpretation. The results supported the following general conclusions:

(1)

Ca²⁺ and Fe²⁺ were found to suppress struvite crystallization through qualitatively different pathways. Ca²⁺ interference was dominated by bulk-phase chemical competition for phosphate and the promotion of Ca-P precipitation routes, which reduces the effective supersaturation available for struvite. In contrast, Fe²⁺ interference was better explained by surface-mediated processes under reducing conditions, where interfacial coordination/adsorption can impede nucleation and growth even without forming detectable crystalline Fe–P impurities. This distinction indicated that the inhibition is not a single phenomenon but a mechanism-dependent outcome.

(2)

Ca²⁺ inhibition can be generalized as a thermodynamic/supersaturation-limiting mechanism.When Ca²⁺ increased, phosphate was preferentially diverted into Ca-P pathways, causing a systematic reduction in the struvite driving force. In nucleation terms, this corresponds to a higher nucleation barrier and a larger critical nucleus, which manifested as delayed crystallization, reduced crystallite size, and higher susceptibility to poorly crystalline by-products. Therefore, Ca²⁺-rich matrices should be viewed as phosphate-availability-limited for struvite formation.

(3)

Fe²⁺ inhibition can be generalized as an interfacial/kinetic mechanism. Under stable anaerobic conditions, Fe²⁺ remained the dominant dissolved iron species and interact strongly with phosphate-containing moieties at the crystal–solution interface. Such interfacial Fe-O-P coordination was consistent with a growth-site blocking mechanism, in which ordered attachment of struvite growth units becomes less favorable. Consequently, Fe²⁺-rich reduced matrices may exhibit suppressed struvite crystallization even when bulk thermodynamics alone would haved predicted favorable precipitation.

(4)

Process control should be mechanism-specific rather than ion-specific. Because Ca²⁺ acted mainly through bulk competition, mitigation should focus on restoring struviting supersaturation and reducing Ca-P diversion (e.g., softening, pH/alkalinity management, staged dosing of Mg²⁺, or controlling phosphate availability). Because Fe²⁺ acted mainly through interfacial inhibition, control should emphasize surface chemistry and adsorption management (e.g., competitive complexants/ligands, surface passivation approaches, or operational strategies that reduce Fe–phosphate interfacial interactions). Treating both ions with the same mindset may lead to inefficient or unnecessary pretreatment.

(5)

A comparative mechanistic framework was found to be essential for complex wastewaters where multiple ions coexist. Real wastewaters often contain Ca²⁺ and Fe²⁺ simultaneously, and their combined effects may not be a linear superposition. Bulk competition by Ca²⁺ may reduce phosphate availability, while surface poisoning by Fe²⁺ may further hinder nucleation or growth, potentially leading to stronger-than-expected suppression. Future studies should therefore evaluate multi-ion coupling effects and develop predictive models that integrate thermodynamic speciation with interfacial kinetics to enable robust process optimization.

The mechanistic interpretations presented here were most applicable to heterogeneous nucleation regimes and the concentration ranges investigated. Extending the framework to full-scale systems will require consideration of additional matrix components and hydrodynamic conditions. Nevertheless, distinguishing bulk-phase competition from surface-controlled inhibition provides a transferable conceptual basis for diagnosing interference types and selecting targeted mitigation strategies.