2.1 Materials

The FeSO₄-ZnSO₄ simulation solution for this test was formulated to replicate the chemical composition of industrial leaching materials, with Zn²⁺ at 140 g/L, Fe²⁺ at 10 g/L, and Cu²⁺ at 0.05 g/L. The chemical composition of the industrial leaching solution from Shenzhen Zhongjin Lingnan Nonfemet Co., Ltd. is shown in Table 1. The neutralizing agent was prepared using deionized water and zinc oxide at a liquid-solid ratio of 9 mL:1 g. Continuous stirring was implemented during usage to ensure the even distribution of its concentration. All reagents used in the experiment were of analytical purity and sourced from Aladdin (China), including FeSO₄·7H₂O, ZnSO₄·7H₂O, CuSO₄·5H₂O, ZnO, and pH buffers solutions. Oxygen with a purity exceeding 99.9% was utilized, and the deionized water used in the experiment was prepared using an ultrapure water instrument. Goethite seeds were prepared through oxidation in FeSO₄ solution at pH 3.5 and 80 ℃ [29].

2.2 Experimental process

All iron removal experiments in this investigation were conducted within a 1 L three-necked flask with a reaction volume of 500 mL. Prior to commencing the experiment, the solution was kept at 80 ℃ using a constant temperature water bath (DF-101S, Shanghai Pailan Instrument Equipment Co., Ltd., China), and its uniformity was ensured through electromagnetic stirring at a rate of 40 r/min. The pH meter (pHSJ-6L, Shanghai INESA Scientific Instrument Co., Ltd., China) underwent calibration using buffer solutions with pH values of 2.0±0.02 and 5.0±0.01. Subsequently, the probe was cleansed with deionized water, and the real-time pH was recorded. Once the solution reaches a stable temperature and pH, the oxygen valve (0.6 L/min) is opened to promote the dispersion and flow of oxygen in a ring of eight 1-mm holes uniformly arranged. The generation of H⁺ ions (Eq. (2)) was continuous throughout the iron removal process. Therefore, to maintain the pH within an error range of ±0.1, an automatic feedback mechanism was employed to supplement the neutralizer (ZnO). The schematic diagram of the iron removal test equipment is shown in Figure 1.

Figure 1全屏查看

Schematic of iron removal devices for goethite process (1-High-pressure oxygen cylinder; 2-Thermostat water bath; 3-Iron removal reactor; 4-pH automatic control system; 5-Neutralizer slurry)

The crystal growth experiment was conducted within a synthetic solution system, with reaction conditions of a pH of 3.5±0.05, temperature of 80 ℃, and reaction duration of 1.5 h. In the heterogeneous nucleation experiment, 0.5 g of crystal seeds were introduced before initiating the iron removal. During the experiment, iron removal solution samples (50 mL) were extracted using a pipette at intervals of 5, 10, 15, 25, 35, 45, 60 and 90 min. These samples were promptly filtered, and the filtrate was reintroduced into the reaction system to maintain a consistent reaction scale. All residue samples were dried in a 70 ℃ oven for 48 h and sealed in vacuum bags for storage. To examine the distribution characteristics of zinc, 10% HCl was employed for the gradual dissolution of the end residues resulting from the heterogeneous reaction. Samples were collected at various dissolution times (2, 10, 20, 40, 60 min) and rapidly filtered to obtain the filtrate. In the case of industrial solutions, iron removal experiments using homogeneous and heterogeneous nucleation were conducted under different pH conditions ranging from 2.5 to 4.0. The filtered residue samples were dried.

2.3 Characterization of samples

Residue phase identification was carried out using an X-ray diffractometer (XRD). The morphology and functional groups associated with crystal growth were observed using scanning electron microscopy (SEM), while Fourier transform infrared spectroscopy (FT-IR) was utilized for analysis. The chemical compositions of all samples were determined using inductively coupled plasma emission spectrometry (ICP-AES). Furthermore, transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS) were employed to analyze the occurrence characteristics of elements within the residue.

3.1 Iron removal rate and residues composition

In Figure 2(a), during the period of 0 to 90 min after iron removal, the concentration of Fe²⁺ in homogeneous and heterogeneous nucleation iron removal solutions decreased from 10 g/L to 0.44 g/L (iron removal rate of 95.6%) and 0.14 g/L (iron removal rate of 98.6%), respectively. Introducing of acicular goethite seeds into the iron removal process has been shown to increase the rate of iron precipitation and purification. HOVE et al [30] and LIU et al [31] concluded that solid carriers providing heterogeneous nucleation are excellent adsorbents for ferrous ions, leading to catalytic oxidation of ferrous ions faster than direct oxidation of ferrous ions in solution. Regrettably, our results do not provide direct evidence to determine whether this process is solely due to the adsorption of crystalline species or accelerated ferrous oxidation via catalytic oxidation.

Figure 2全屏查看

(a) Concentration of Fe²⁺ in zinc-rich solution; The content of (b) Fe and (c) Zn in the residues with time under two different nucleation modes at 80 ℃, with pH of 3.5±0.1

The theoretical iron content of goethite crystals is 62.92%. However, the iron content of the residues was lower than the theoretical value due to the low crystallinity and impurities in the precipitates. Therefore, achieving high crystallinity and minimizing impurity phases is vital to obtaining high-content residues. In the homogeneous nucleation process, there was a significant variation in the iron content of the residues over time (Figure 2(b)). During the initial stage (0-15 min), the iron content of the precipitates ranged from 45.15% to 47.58%, indicating that the initial residues were primarily composed of hydrolyzed polymers with high water content and amorphous crystal nuclei. The iron content of the residues increased with time, reaching 56.22% at 35 min.

Comparatively, the addition of seed increased the iron content of the initial residue (at 15 min) by 24.18%, which indicates that the Fe³⁺ induced by seed may directly contribute to the secondary growth of the nucleus, leading to the formation of high-crystallinity goethite crystals [32]. However, at 35 min, the iron content of the residues declined to 58.67%, indicating that apart from participating in crystal growth, another portion of Fe³⁺ started forming new crystal nuclei. The iron content of the residues in heterogeneous nucleation remained higher, measuring 55.67% (without seed addition) and 59.23% (with seed addition) in the final residues.

The loss of zinc in precipitates impacts the yield of the final product and can lead to an increase in the production of residues. Thus, the zinc content in residues at different times was examined to identify its loss pattern, as shown in Figure 2(c). Across all samples, the residue without seed addition consistently exhibited higher zinc content than those with seed addition. Meanwhile, the zinc content in the homogeneous nucleation decreased from 9.76% to 7.58% with increasing time. In contrast, the heterogeneous nucleation displayed an increasing trend of zinc content, rising from 4.11% to 5.34% over time. This result suggests that the crystallinity of the residues influences the amount of zinc loss.

3.2 Structural and microtopography of goethite growth

3.2.1 XRD and FT-IR

In industrial production, the crystallinity of iron precipitates plays a significant role in determining the number of residues, zinc loss, and filtration performance. Hence, XRD and FT-IR were used to analyze the phase and molecular structure of residues under two nucleation methods to understand the evolution of goethite crystal (Figure 3).

Figure 3全屏查看

XRD (a, c) and FT-IR (b, d) of goethite residues with time: (a, b) Homogeneous nucleation; (c, d) Heterogeneous nucleation

In the absence of goethite seeds, the characteristic peak of goethite (α-FeOOH) characteristic peak could not be observed in the diffraction pattern within the initial 15 min, indicating that the residues were predominantly amorphous (Figure 3(a)). In particular, the pH is often challenging to regulate stably during industrial production, thus easily leading to colloid formation or excessive nucleation during the nucleation stage, which is detrimental to crystal growth. At 25 min, the presence of characteristic peaks of goethite indicated the reaction’s transition into the crystal growth stage. For 35 to 90 min, signifying an enhancement in the crystallinity of the residue as the crystal continued to develop in the growth stage [33]. Notably, when compared to the PDF#99-0055 standard card, the prominent crystal faces (110), (130), (021) and (111) exhibited normal development, but the growth of crystal faces (120) and (041) was inhibited, which might be influenced by the high concentration of ZnSO₄.

Figure 3(c) displays the XRD analysis of goethite residue during the heterogeneous nucleation process. With the addition of seeds, the initial residue exhibited high crystallinity, which indicates that the induction of seeds effectively prevented the formation of a large amount of amorphous residue during the early stages of iron removal. Unlike homogeneous nucleation, the intensity of the characteristic peaks in heterogeneous nucleation does not consistently increase with time. Between 0 and 25 min, the peak values of (120) crystal planes increased, while the peaks of (020), (110) and (130) slightly decreased. This suggests that although the crystal planes of the seeds continued to grow, the development speed of each crystal plane varied. After 35 min, the crystallinity of the residue decreased as the initially developed goethite crystals reached their size limit. At this stage, the rate of Fe³⁺ consumption for crystal growth became slower than the rate of Fe²⁺ oxidation to Fe³⁺. As a result, some of the Fe³⁺ participated in forming new nuclei, leading to a coexistence of crystal growth and the formation of new nuclei. Moreover, the addition of crystalline seeds maintained a heterogeneous induction throughout the system, with only a small portion of the partially oxidized Fe³⁺ involved in forming new nuclei. This reduced the susceptibility to burst nucleation phenomena.

Figures 3(b) and (d) present the FT-IR spectra of goethite residues obtained from two nucleation modes. The α-FeOOH absorption peaks were observed at different times. The absorption peak at 3418 cm^-1 corresponds to the expansion vibration of adsorbed H₂O, while the peak at 3158 cm^-1 represents the O—H stretching vibration in goethite. The 897 cm^-1 and 792 cm^-1 bands correspond to the prominent Fe—O—H bending vibrations. In addition, the peak at 1134 cm^-1 corresponds to the bending vibration of SO₄^2-. During homogeneous nucleation, the vibrations of both O—H and Fe—O—H increased gradually over time (Figure 3(b)). Before 25 min, the bending vibration of the goethite absorption peak was relatively weak. EUSTERHUES et al [34] believed that this is related to the crystallinity of goethite, which is also consistent with the low crystallinity results indicated by XRD at the present stage. In the experiment with the addition of seeds, the vibrational peaks of —OH at 3158 and 1637 cm^-1 were observed at 5 min, which implies that FeOOH was the main phase of the precipitates in the early stage of the iron removal (Figure 3(d)). The results show that the iron residue produced at the initial stage of homogeneous nucleation has a low acicular ferrite character, while the product obtained by heterogeneous nucleation can always maintain a high degree of crystal crystallinity. Furthermore, the absorption peak at 1025 cm^-1, corresponding to the Fe—OH vibration of β-FeOOH, was observed, suggesting the presence of β-FeOOH entrapped in the seeds. The vibrational intensity of the absorption peak decreased over time, indicating that no new lepidocrocite was precipitated during the iron removal process.

3.2.2 SEM and TEM

The morphology of the residues with time was observed by scanning electron microscopy (Figure 4), and the projection images of the endpoint residues were analyzed by transmission electrolysis (Figure 5).

Figure 4全屏查看

SEM images of goethite residues at different times: (a) Homogeneous nucleation; (b) Heterogeneous nucleation

Figure 5全屏查看

TEM images of (a) homogeneous and (b) heterogeneous nucleation of goethite residues at 90 min

Without the addition of seeds, the entire solution system was initially homogeneous, and the Fe³⁺ generated through oxidation would undergo hydrolysis to form polymorphs and crystal nuclei [35]. As shown in Figure 4(a), at 5 min, the microscopic morphology primarily consisted of closely connected spherical particles with diameters ranging from 5 to 10 nm. Although some nuclei began developing into goethite crystals during the continuous iron removal process, small particles were still observable, implying that the nucleation process continued until 35 min. It also shows that crystal growth and nucleation occur simultaneously in the early stages of iron deposition. It may be attributed to the higher oxidation rate of Fe²⁺ to Fe³⁺ compared to the rate of crystal growth consumption, which made the excess Fe³⁺ participate in forming nuclei [36]. By 45 min, the residue morphology predominantly comprised needle-like crystals, which shows that the nuclei were no longer generated at this time, and Fe³⁺ was mainly committed to crystal growth.

In the case of heterogeneous nucleation within the 5-20 min, only needle-like crystals were observed without any spherical particles, indicating that the addition of seeds directed Fe³⁺ to preferentially participate in the crystal growth process rather than nucleation formation (Figure 4(b)). Therefore, goethite seeds provided an abundant adsorption interface for Fex⁺ (x=2, 3), thereby accelerating the iron removal rate compared to the homogeneous nucleation process. Also, the seeds helped prevent the excessive production of crystal nuclei during the early stages. Notably, after 35 min, the presence of amorphous particles and fine needle-like crystals was observed, indicating that nucleation was inevitable but delayed, and the number was significantly smaller. Between 60 and 90 min, the crystals appeared in strip-like formations, with the initially formed crystals covered within the agglomerates.

The residues obtained after 90 min of reaction under the two nucleation methods were analyzed using TEM (Figure 5). In the high-magnification images, defects on the crystal surface can be observed in the crystals obtained without adding seeds (Figure 5(a)). On the other hand, the crystals obtained through heterogeneous nucleation exhibit a smooth and complete surface (Figure 5(b)). The residues obtained by the two methods can eventually form needle-like crystals, but the crystals obtained by heterogeneous nucleation have larger average sizes and higher crystallinity than those obtained through homogeneous nucleation.

3.3 Existence mechanism of zinc in residues

The Fe, Zn and S elements in the residue were analyzed using TEM surface and profile scans (Figure 6). Meanwhile, the distribution relationship between Zn and Fe in the residue was investigated by gradually dissolving with dilute acid (Figure 7).

Figure 6全屏查看

TEM mapping and profile scan of (a, c) homogeneous and (b and d) heterogeneous nucleation of goethite residue (HAADF: high-angle annular dark field)

Figure 7全屏查看

Result of pickling of goethite residue: (a) Zn²⁺ and Fe³⁺ concentrations in solution; (b) ΔC_Zn/ΔC_Fe ratios at different time intervals

In both nucleation methods, the elements Fe, Zn, and S exhibited a uniform distribution on the goethite crystals without any concentrated regional patterns, and their elemental content decreased successively (Figures 6(a) and (b)). The crystals obtained through homogeneous nucleation had Zn and S contents of 6.78% and 0.84%, respectively, corresponding to an atomic ratio of 4:1. It exceeded the theoretical value of 1:1 for Zn:S in ZnSO₄, confirming that Zn deposition was not solely attributed to ZnSO₄ adsorption. In addition, based on the crystal profile scan results, the S content showed no significant variation with crystal thickness, indicating that S was primarily adsorbed on the crystal surface and did not penetrate in the interior of the goethite crystals during growth (Figures 6(c) and (d)). Meanwhile, the Zn content tended to increase with increasing thickness at different scanning positions, suggesting that, in addition to surface adsorption, Zn also participated in the crystal lattice during the goethite development process [37]. This observation aligns with LI et al’s view that zinc in waste residues predominantly exists through surface adsorption and crystal embedding mechanisms [38]. Notably, in the zinc recovery process, the surface adsorbed zinc could be recovered by washing with hot water or dilute acids, while crystal-embedded zinc required extraction methods such as roasting or pressure leaching [6, 39]. Although the current industrial residues have recycling value, the associated processes are energy-intensive and environmentally polluting [9]. Therefore, minimizing sediment formation and reducing zinc loss is crucial for cost-effective residue recovery.

The content of Fe and Zn ions at 2, 10, 20, 40 and 60 min of dissolution was analyzed using the hydrochloric acid dissolution method (Figure 7(a)). After 2 min, the Fe and Zn contents were measured as 0.27 g/L and 0.11 g/L, respectively. The zinc adsorbed on the surface of the residue was released into the solution, which is a crucial aspect of the washing recovery process. As the dissolution progressed, the Fe and Zn contents continued to increase. By 60 min, the Fe concentration reached 2.69 g/L (10 times that at 2 min), while the Zn concentration was only 0.29 g/L (2.64 times that at 2 min). The rate of increase in Zn concentration was much lower than that of Fe concentration. Thus, we calculated the Zn to Fe ion increment ratio for different time intervals (Figure 7(b)). In the 0-2 min interval, ΔC_Zn/ΔC_Fe was 0.42, indicating a substantial release of Zn into the solution during the dissolution of goethite. In the subsequent four intervals (2- 10 min, 10-20 min, 20-40 min, and 40-60 min), the ΔC_Zn/ΔC_Fe ratios ranged from 0.06-0.086, which suggests that the zinc within the crystal is dispersed and uniformly distributed, despite accounting for a significant portion of zinc loss.

XPS was used to ascertain the composition of the elements and the chemical state of the residues’ surfaces. In this study, the heterogeneous nucleation residues and their samples subjected to a 2 min acid washing were subjected to XPS analysis. All spectra were calibrated using Avantage software, with the C 1s peak (284.8 eV) as the reference. The narrow spectrum peaks were accurately fitted and separated. The comprehensive spectral outcomes revealed that the O and Fe peaks exhibited robust signals with negligible variation before and after the acid-washing process, whereas the signal intensity of Zn 2p was significantly diminished (Figure 8(a)). Figure 8(b) shows the residues’ narrow spectrum of Fe 2p. Both Fe 2p_1/2 and Fe 2p_3/2 were separated into two peaks, with binding energies measured at 726.39, 724.48 and 713.59, 711.20 eV, respectively [40]. The binding energies of Fe 2p_1/2 and Fe 2p_3/2 in the acid-washed residues displayed minimal shifts. The O 1s peaks were attributed to Fe—OH and Fe—O, and their binding energies, before and after the acid treatment, were found to be 531.54 eV, 530.07 eV and 531.43 eV, 530.07 eV, respectively, with negligible displacement (Figure 8(c)). Hence, the phase structure of the residues after acid washing was still FeOOH. Notably, the binding energy of Zn 2p_3/2 and Zn 2p_1/2 exhibited a reduction of 0.23 eV after the pickling process compared to the pre-pickling state, accompanied by a decrease in peak intensity (Figure 8(d)). This further corroborates that the existing form of zinc on the surface differed from its internal crystal state, highlighting the concentrated distribution of zinc on the residue’s surface.

Figure 8全屏查看

XPS of goethite residue before and after pickling for 2 min: (a) Wide spectra; (b) Fe 2p; (c) O 1s; (d) Zn 2p

3.4 Analysis of application in industrial materials

Homogeneous and heterogenic tests were carried out with real industrial solutions at different pH conditions. The iron removal temperature was 80 ℃, the oxygen flow rate was 0.6 L/min, the neutralizer was ZnO, and the reaction time was 3 h.

Figure 9 shows that the crystal structure of goethite residues at different pH had apparent changes. At pH 2.5, sodium jarosite (NaFe₃(SO₄)₂(OH)₆) was mainly observed in the homogeneous nucleation residue, and no characteristic peak of goethite (α-FeOOH) was found (Figure 10(a)). This is because the growth of FeOOH is difficult at low pH, and the presence of Na⁺ satisfies the conditions for the precipitation of sodium jarosite [41, 42]. When the pH was controlled at 3.0, sodium jarosite and goethite were observed in the residue, which shows that the pH 3.0 condition is the coexistence interval for the growth of both crystal forms. At pH 3.5, sodium jarosite can not precipitate, and goethite with low crystallinity was the primary phase. At pH 4.0, the hydrolyzed residues encountered difficulty forming crystals and instead precipitated in an amorphous state. SEM results revealed that at pH 2.5, the microscopic morphology of the residues was dominated by lamellar hexagonal sodium jarosite crystals, with a few presence of amorphous agglomerates (Figure 9(b)). Sodium jarosite crystals possess a smaller specific surface area than goethite crystals, resulting in lower zinc adsorption on their surface, which is consistent with the generally low percentage of zinc in sodium jarosite residues [43]. The distinct needle-like crystal form was observed at the optimal pH of 3.5. At higher pH 4.0, the residues were predominantly composed of amorphous agglomerates, which could not form crystals during precipitation.

Figure 9全屏查看

XRD and SEM of goethite residue obtained at different pH in actual industrial solution: (a, b) Homogeneous; (c, d) Heterogeneous nucleation

Figure 10全屏查看

Amount of (a) zinc loss and (b) residue yield at different pH

The introduction of seeds resulted in an enhanced level of crystallinity in the residues obtained under varying pH conditions. This is evident in the XRD pattern, which displays the presence of the α-FeOOH phase in all residues obtained from pH 2.5 to 4.0 (Figure 9(c)). Although the precipitation of sodium jarosite was unavoidable at low pH 2.5 and 3.0, the induction of seeds made the observation of needle-like crystalline forms in the residue at low pH (Figure 9(d)). In particular, at the optimum growth pH of 3.5 for goethite, the average crystal size was more extensive than that of the residue obtained through homogeneous nucleation under the same experimental conditions. Furthermore, the needle-like precipitation was still noticeable at pH 4.0, although the overall crystallinity of the residue was reduced.

According to the content of zinc and iron in the residue, the zinc loss and residue amount for each ton of iron removed was calculated, respectively (Figure 10). The loss of zinc in the residue showed an increasing trend with increasing pH, as shown in Figure 10(a). At pH3.5, the residue had a high crystallinity, and its zinc loss was low. At pH 2.5, the loss of zinc was 109.19 kg/t Fe and 78.03 kg/t Fe for homogeneous and heterogeneous nucleation, respectively. It is clear that the low zinc loss at low pH resulted from the smaller specific surface area of sodium jarosite compared to goethite, and the surface adsorption of zinc was low. At pH 3.5, seed addition reduced zinc loss by 50.91 kg/t Fe. When the pH reached 4.0, the loss of Zn from homogeneous and heterogeneous nucleation increased to 340.32 kg/t Fe and 290.16 kg/t Fe, respectively, which is the contribution of the decrease in residue crystallinity and increase in specific surface area. Furthermore, although no zinc hydrolysis phase was observed through XRD analysis, the pH was close to the hydrolytic precipitation region of zinc, so some zinc may also undergo hydrolytic precipitation.

The residue obtained at different pH conditions decreased and then increased with pH (Figure 10(b)). At pH 2.5, the residue yield for homogeneous and heterogeneous nucleation reached 2.89 t/t Fe and 2.58 t/t Fe, respectively. The high residue yield was attributed to the low iron content of NaFe₃(SO₄)₂(OH)₆ formation. Sodium jarosite was no longer precipitated at pH 3.5, goethite crystals dominated, and residue reached the lowest amount (The residue content was reduced by 0.17 t/t Fe compared to heterogeneous nucleation). Increasing the pH to 4.0 would result in high bound water content in the low crystallinity residue, while the possible hydrolytic precipitation of zinc further increases the residue volume [44]. In conclusion, the amount of zinc loss and residue of the iron removal process with the seed at different pH was lower than that of homogeneous nucleation, which is supposed to be the contribution of higher crystallinity. Therefore, the industrial addition of seed is a vital measure for low zinc loss and residue volume.