Role of Iron Impurity in Hydrometallurgical Recovery Process of Spent Lead-Acid Battery: Phase Transformation of Positive Material Made from Recovered Leady Oxide

The chemical compositions (62.2 wt% PbSO₄, 21.8 wt% PbO₂, 14.2 wt% PbO, and 1.0 wt% metallic Pb) of simulated spent lead paste were designed according to the mineral phases (Figure 1a) and chemical compositions (Table I) of spent lead paste provided by Hubei Jinyang metallurgical Co., Ltd. As shown in Figure 1b, Fe₂O₃ and Fe₃O₄ are identified as the major mineral phases of the magnetically concentrated lead paste sample. Hence, the mixture of Fe₂O₃ and Fe₃O₄ (molar ratio of Fe₂O₃ to Fe₃O₄ was designed as 1:1) was mechanically mixed with the simulated spent lead paste to prepare the Fe-containing simulated spent lead paste sample. Different mass ratios of total Fe element to the simulated spent lead paste were set as 0, 0.002, 0.005, 0.02, 0.05, 0.2, and 0.5 wt%, respectively.

Zoom In Zoom Out Reset image size

According to our previous studies,¹⁹ 100 g Fe-containing simulated spent lead paste was added to a solution containing 163.2 g sodium citrate (Na₃C₆H₅O₇·2H₂O, purity ≥ 99% W/W), 35 mL acetic acid (CH₃COOH, purity ≥ 99.5% W/W), and 60 mL hydrogen peroxide (H₂O₂, purity ∼30% W/V) with a solid/liquid mass ratio of 1/8. The slurry was stirred using an overhead stirrer at room temperature (25°C) with a speed of 300 rpm for 6 h. Subsequently, the mixture was kept at 50°C of 24 h for the growth of lead citrate crystals. The lead citrate obtained after filtration and washing was dried at 50°C for 5 h. Finally, the lead citrate was calcined at 375°C in the air to synthesize the Fe-containing leady oxide. The leady oxide samples prepared from the simulated spent lead pastes with the Fe element contents of 0, 0.002, 0.005, 0.02, 0.05, 0.2, and 0.5 wt% were referred to as LO-0, LO-0.002, LO-0.005, LO-0.02, LO-0.05, LO-0.2, and LO-0.5, respectively. The residual ratio of Fe element during the whole hydrometallurgical process was calculated according to Eq. 6.

where m₁ (g) is the mass of the obtained leady oxide; W₁ (wt%) is the mass percentage of Fe element in the obtained leady oxide; m₂ (g) is the mass of the simulated spent lead paste; and W₂ (wt%) is the mass percentage of Fe element in the simulated spent lead paste.

The Fe-containing leady oxide was mechanically mixed with 0.1 wt% fiber and 14 wt% deionized water, followed by slowly adding 9.0 wt% H₂SO₄ solution (1.4 g cm⁻³) for the paste mixing. The apparent density of paste was higher than 4.0 g cm⁻³ after mixing. The obtained wet lead paste of 27.5 g was pasted on a Pb-Ca-Sn-Al alloy plate (35 mm × 60 mm × 1.2 mm) to manufacture a 2 Ah positive plate based on the PAM utilization efficiency of 45%.

The curing of the positive plate was carried out in a curing vessel under the following conditions: 80°C and 95% relative humidity for 48 h, followed by 70°C and 10% relative humidity for 24 h. In the formation procedure, the cured positive plate was soaked in a sulfuric acid solution (1.05 g cm⁻³) for 2 h, followed by the constant current charged for 48 h (triple theoretical capacity). Each positive plate was coupled with two commercial negative plates and followed by adding 40 mL sulfuric acid (1.24 g cm⁻³) for a 2 V/2 Ah testing cell assembly.

The initial discharge capacities (C/20 and 1C) of the testing cells were measured at the discharge current of 0.1 A and 2 A, respectively. The charge-discharge cycling test was carried out under the following process: during charging the cell was first charged at 0.2 A (C/10) until the cell voltage increased up to 2.45 V and followed by constant voltage charging at 2.45 V for 6 h, then the 100% depth of discharge (100% DOD) of cell was performed at 0.2 A (C/10) with the cut-off potential of 1.75 V. All the cell cycling tests were carried out under constant temperature of 25±1°C. The cell was considered to be a failure state when the discharge capacity decreased to 1 Ah.

The content of Fe element in the leady oxide samples was measured by an inductive coupled plasma optical emission spectrometer (ICP-OES 8300, PerkinElmer, USA) after digesting the samples with 5% V/V HNO₃ solution. The X-ray diffraction (XRD) analysis was carried out using X-ray diffraction analyzer (XRD-7000, Shimadzu, Japan) with Cu Kα radiation in the range from 5° to 75°. The X-ray photoelectron spectroscopy (XPS) was employed for the measurement of hydrated PbO₂ content in the PAMs with Al-Kα X-ray source operating at 10 mA and 15 kV (AXIS-ULTRA DLD-600W, Shimadzu, Japan). The specific surface and pore size of the PAMs after formation were measured using a BET surface area analyzer (BK112-TB, JWGB Sci & Tech. Co., Ltd., China) based on nitrogen adsorption and desorption method. The morphology of samples was studied using a field emission scanning electron microscope (FESEM, Nova Nano SEM 450, FEI, USA). The high-resolution transmission electron microscope image of hydrated and crystalline PbO₂ was obtained with a transmission electron microscope (TEM, Talos F200x, FEI, USA) after the PAM was manually shattered and dispersed by ultrasound in ethanol. The microanalyses of Pb, O, S, and Fe elements in the samples were carried out by energy dispersive X-ray spectroscopy (EDX) and high-resolution hyper probe wavelength-dispersive spectroscopy (WDS) with an electron probe micro-analyzer (EPMA, 8050G, Shimadzu, Japan).

The contents of crystalline phases in the positive material were expressed by the relative intensities of the characteristic diffraction lines in the XRD patterns, I_a/ΣI_n, where I_a represents the characteristic diffraction line for a given phase and ΣI_n is the sum of the intensities of the characteristic lines for all crystalline phases in the samples.^21,22 The following characteristic reflections were used: d = 0.307 nm for α-PbO, d = 0.295 nm for β-PbO, d = 0.322 nm for 4BS, d = 0.326 nm for 3BS, d = 0.313 nm for α-PbO₂, and d = 0.350 nm for β-PbO₂.

As shown in Table II, the residual ratios of Fe element in prepared leady oxide samples were all more than 75 wt% with various Fe additions, indicating that most of the Fe element impurity in the simulated spent lead paste remained in the leady oxide after the leaching and calcination process. In addition, the recovery efficiency of Pb were all more than 98 wt%.

This result can be explained by the fraction diagrams of Pb²⁺ and Fe³⁺ in Fe³⁺-Pb²⁺-cit^3–-CH₃COO⁻-SO₄^2– system (cit^3– for citrate ion, C₆H₅O₇^3–) constructed by Medusa software (Figures 2a and 2b). In this sodium citrate-acetic acid solution leaching system, the pH of leaching solution was controlled within a range of 5.0–6.0 to achieve a high recovery efficiency of Pb element.¹⁹ Figure 2a shows that the prepared lead citrate precursor was mainly Pb₂(cit)₃ under such condition. This result is consistent with previous study that the product was Pb₃(C₆H₅O₇)₂·3H₂O when the leaching pH was 5.2.²³ Figure 2b shows that the Fe³⁺ ions were formed into solid phase of Fe₂O₃ when the pH was 5.0–6.0. Therefore, the Fe₂O₃ remained in the lead citrate precursor after filtration, and then incorporated into the final leady oxide products after calcination of lead citrate precursor. The results of the contents of PbO in different leady oxide samples indicate that Fe impurity has no obvious effect on the oxidation degree of leady oxide in the lead citrate calcination step (Table II).

Zoom In Zoom Out Reset image size

The transformations of Pb and Fe elements in the hydrometallurgical recovery of simulated spent lead paste with Fe element content of 0.05 wt% are shown in Figure 2c as an example. As shown in Figure 2c, 98.5% of Pb element in the simulated spent lead paste was recovered after leaching step, and 86.6% of Fe element in the simulated spent lead paste remained in the lead citrate precursor. After the calcination step, most of the Pb and Fe elements in the lead citrate precursor remained in the leady oxide product. The total recovery efficiency of Pb element in the whole recovery process was calculated as 98.4%, and the total residual ratio of Fe was 86.2%. The mass balances of Fe element during the recovery of other Fe-containing simulated spent lead paste samples are also calculated and provided in Table III.

In the curing procedure, some PbO in leady oxide could react with H₂SO₄ to generate the intermediate phase of 3BS. Then the 3BS would react with PbO to generate 4BS under higher temperature when a certain ratio of PbO to 3BS was achieved.²⁴ The major crystalline phases of the CPPs are identified as 3BS, 4BS, and β-PbO, as shown in Figure 3a. The changes in relative intensity of different phases are presented in Figure 3b. When the content of Fe in the leady oxide samples was lower than or equal to 223 mg kg⁻¹ (LO-0, LO-0.002, LO-0.005, and LO-0.02), the relative intensities of all phases in the prepared CPPs remained basically unchanged. However, the relative intensity of 4BS in the XRD patterns of the CPPs decreased significantly as the content of Fe in the leady oxide continued to increase up to 540 mg kg⁻¹ (LO-0.05), while the relative intensity of 3BS increases accordingly. In addition, some of the β-PbO did not convert to lead sulfate and remained in the CPPs, and its relative intensity almost kept constant with different Fe contents. When LO-0.5 (Fe content of 5129 mg kg⁻¹) was used, the relative intensities of 4BS, 3BS, and β-PbO in the XRD pattern of the CPP were 0, 69.6, and 30.4%, respectively, indicating that no 4BS could be identified under such high content of Fe impurities.

Zoom In Zoom Out Reset image size

Figure 4a represents the SEM images of the CPP made from LO-0. It consisted of rod-like crystals within the length range of 50–100 μm. The atomic ratio of S:Pb:O element of the selected zone was 1:4.96:8.55, as presented in the EDX result (Figure 4b). It indicates that these rod-like crystals were 4BS crystals (1:5:8), which is consistent with the results of the previous studies.^25,26 These large-size 4BS crystals can generate a durable positive plate strength after formation, contributing to a longer cell charge-discharge cycle life. Figure 4c presents the back-scattered electron (BSE) image of the CPP made from LO-0.2 by EPMA. It shows that the sample contains only a small amount of rod-like crystals. The WDS analysis result (Table IV) of the selected rod-like crystals at Position 1 shows that the atomic ratio of S:Pb:O was 1:4.78:7.39:7.74, which was close to the composition ratio of 4BS, indicating that these rod-like crystals were also 4BS. By contrast, the atomic ratio of S:Pb:O at Position 2 and Position 3 were 1:102.5:97.5 and 1:3.68:7.35, respectively, which was close to the composition ratio of PbO (0:1:1) and 3BS (1:4:7), indicating that the major phase was PbO and 3BS at Position 2 and Position 3, respectively. The corresponding WDS mapping result of Fe element is shown in Figure 4d. The darker color represents the lower content of Fe element. Based on the distribution of Fe element, this WDS mapping image can be roughly divided into two areas. The upper right and lower left zones represent the higher-Fe area, and the middle zone represents the lower-Fe area. Combine with the BSE image in Figure 4c and EPMA-WDS mapping result in Figure 4d, it can be concluded that larger 4BS crystals were mainly distributed in the lower-Fe area. This is also consistent with the result in Table IV that the contents of Fe element at Positions 1, 2, and 3 are 0.05, 0.74, and 0.81 wt%, respectively. Therefore, the above results indicate that the generation of 4BS in the CPP was inhibited by Fe element impurities.

Zoom In Zoom Out Reset image size

From the above analysis, the Fe element impurity has a significant influence on the 3BS/4BS ratio of the CPPs. When the Fe element content in leady oxide was higher than 223 mg kg⁻¹ of LO-0.02, the generation and crystallization of 4BS were significantly inhibited.

Figure 5a shows XRD patterns of PAMs after formation procedure. The diffractograms for all samples demonstrate the presence of α-PbO₂ and β-PbO₂ phases. The intensity of the strongest crystalline peak (27.8^o) of α-PbO₂ became weaker with the increasing of Fe content. Figure 5b shows that the relative intensities of α-PbO₂ and β-PbO₂ phases in the PAMs after formation procedure remained basically unchanged when the content of Fe in the leady oxide was lower than or equal to 223 mg kg⁻¹ of the leady oxide samples (LO-0, LO-0.002, LO-0.005, and LO-0.02). The relative intensity of α-PbO₂ decreased significantly as the content of Fe increased up to 540 mg kg⁻¹ of LO-0.05, while the relative intensity of β-PbO₂ increased accordingly. For the PAM made from the LO-0.5, the relative intensities of α-PbO₂ and β-PbO₂ are 7.6 and 92.4%, respectively. According to the potential-pH diagram for the Pb-H₂O-H₂SO₄ system constructed by Ruetschi,²⁷ α-PbO₂ was usually generated in alkaline environment, while β-PbO₂ was generated in acidic environment. During the formation procedure, the 4BS phase generates more OH⁻ ions than 3BS phase (Eqs. 1–2). In addition, the PbO₂ formed on the surface of the large 4BS crystals was always denser than that on the small 3BS crystals.²⁸ As a result, the diffusion of H₂SO₄ electrolyte into 4BS crystals is more difficult than 3BS crystals, so the pH of the electrolyte inside the 4BS-based plate is more alkaline, resulting in the generation of more α-PbO₂. The α-PbO₂ phase was more favorable for cell life because it has a tighter structure than β-PbO₂, which was beneficial to prevent the PAM shedding during the charge-discharge cycling.^29,30

Zoom In Zoom Out Reset image size

Figure 6a shows the low-resolution SEM image of the PAM made from LO-0. It can be seen that the PAM maintains the skeleton structure of 4BS. The high-resolution SEM image in Figure 6b shows the presence of a large amount of tightly bound small particles in PAM. These microporous, smaller particles were fairly well interconnected into an electron conductive and mechanically stable skeleton for PAM. The EDX analysis in Figure 6c indicates that these particles were PbO₂ since the atomic ratio of Pb:O was 1:1.93. Figure 6d shows the low-resolution SEM image of the PAM made from LO-0.2. It can be seen that the PAM has no obvious skeleton structure. The high-resolution SEM image in Figure 6e shows that the particles generated PAM made from LO-0.2 were significantly larger and looser distributed than those made from LO-0. The EDX analysis in Figure 6f indicates that these large particles were also PbO₂ since the atomic ratio of Pb:O was 1:2.20. Previous studies have noted the importance of more microporous and smaller PbO₂ structure in the cycling performance of cells.¹² During the charge-discharge cycling, larger PbO₂ was unfavorable to the electrochemical and chemical reactions of the positive plates due to the extremely poor contact between the larger PbO₂ particles and electrolyte, i.e., the H⁺ and HSO₄⁻ ions.¹² As a result, the discharge capacity of the positive plate with larger PbO₂ particle was lower. In addition, the dense structure in PAM made from LO-0 is beneficial for high conductivity and mechanical strength of the plate.⁸

Zoom In Zoom Out Reset image size

Figure 7a shows the pore volume distribution vs. the pore diameter of the PAMs after formation procedure. Generally, the electrochemical processes of cell discharge take place on the surfaces of these small pores with a diameter smaller than 100 nm. The PAMs made from lower Fe content leady oxide samples (LO-0, LO-0.002, LO-0.005, and LO-0.02) showed significantly larger pore volume than PAMs with higher Fe content (LO-0.05, LO-0.2, and LO-0.5). Furthermore, Figure 7b shows that the BET surface area of the PAMs made from LO-0.05, LO-0.2, and LO-0.5 is obviously smaller than those made from LO-0, LO-0.002, LO-0.005, and LO-0.02. Larger pore volume and higher surface area are beneficial for facile mass transfer of H₂SO₄ electrolyte during the cell charge-discharge cycling.

Zoom In Zoom Out Reset image size

Figure 8a shows the TEM image of gel/crystal zones of PbO₂ in the PAM made from LO-0. The dark zones are the crystals of a-PbO₂ or β-PbO₂, while the lighter regions indicate gel zones of hydrated PbO₂.³¹ The EDX analysis in Figure 8b shows that the selected zone was composed of PbO₂ since the atomic ratio of Pb:O:S was 1:2.15:0. In the high-resolution TEM image of the selected zone (Figure 8c), some clear crystal zones can be observed from the composite. The crystal in the upper left zone reveals an interlunar distance of 0.313 nm, which agrees with the interlunar distance of (1, 1, 1) crystal plane of α-PbO₂. The crystal in the bottom right zone is identified as β-PbO₂ due to its interplanar distance of 0.350 nm. In addition, the PAM contained a lot of hydrated PbO₂ in the upper right zone. Figure 8d shows the TEM image of gel/crystal zones of PbO₂ in the PAM made from LO-0.2. It shows that the PAM contains more dark zones than the PAM made from LO-0. The EDX analysis in Figure 8e shows that the selected zone was composed of PbO₂ since the atomic ratio of Pb:O:S was 1:2.62:0. The high-resolution TEM image (Figure 8f) reveals the presence of big β-PbO₂ crystal in PAM.

Zoom In Zoom Out Reset image size

To further investigate the hydrated PbO₂ content in the PAMs after formation procedure, XPS spectra of PAMs were acquired. The binding energy was calibrated with respect to the C 1s peak at 284.6 eV. The full spectrum of PAM made from LO-0 after formation in Figure 9a reveals the presence of Pb and O elements. The deconvoluted XPS spectra of O ls peak (Figure 9b) disclose the presence of three peaks at 531.5±0.2 eV (S-O from PbSO₄), 530.2±0.2 eV (O-H from hydrated PbO₂), and 529.3±0.2 eV (Pb-O from PbO₂).^12,32 As shown in Figure 9c, the peak area of O-H significantly decreased after the discharge ending of 3^rd cycle, owing to the transformation from PbO₂ into PbSO₄. It indicates that the hydrated PbO₂ in gel zones was converted into PbSO₄ during discharge. When the plate was re-charged, the gel zone was re-established.¹¹

Zoom In Zoom Out Reset image size

The ratio of the peak area of O-H to the sum of the peak areas of O-H and Pb-O, defined as S_O-H/(S_O-H + S_Pb-O), is used to present the relative content of hydrated PbO₂ in the PAMs.¹¹ Figure 9d shows the hydrated PbO₂ content of the PAMs with different Fe contents. When the Fe content of the leady oxide was lower than or equal to 223 mg kg⁻¹ of leady oxide samples (LO-0, LO-0.002, LO-0.005, and LO-0.02), the hydrated PbO₂ content remained stable, while it decreased significantly when the Fe content of the leady oxide was higher than 540 mg kg⁻¹ in LO-0.05. After the discharge ending of 3^rd cycle, the hydrated PbO₂ content of the PAM also showed a decreasing trend when the Fe content of the leady oxide was higher than 223 mg kg⁻¹.

Figure 10a shows the BSE image of the PAM made from LO-0 after the discharge ending of 3^rd cycle. The WDS mapping results (Figures 10b and 10c) show that the larger crystal with the size of about 20 μm is PbSO₄, and the small agglomerated crystals are PbO₂. The nanostructured PbO₂ particles adhered to the surface of the PbSO₄ crystals. In the re-charge step, these PbO₂ particles acted as conductive units to promote the conversion of PbSO₄.³³ As for the PAM made from LO-0.2, the SEM image shown in Figure 10d indicates that the sample consisted of particles with a large size of 30–80 μm. The EDX analysis of the selected zone (Figure 10e) indicates that the crystals were PbSO₄ particles since the atomic ratio of Pb:S:O was 1:0.99:4.46. Due to the large particle size of PbSO₄, the access from the exterior H₂SO₄ electrolyte to interior PbSO₄ crystals was hindered, and the charge acceptance of the plate was significantly deteriorated.³⁴

Zoom In Zoom Out Reset image size

The average crystallite size of PbSO₄ was calculated from the XRD patterns using the Scherrer equation.^35,36 Figure 10f shows the variation of the average crystallite size of PbSO₄ in the PAMs after the discharge ending of 3^rd cycle. It can be seen that the PbSO₄ crystallite size in the PAMs was increased with the increase of Fe content in the PAMs. Also, those PAMs with higher Fe impurity content contained more PbSO₄. This result is also consistent with the SEM images in Figures 10a and 10d.

Figure 11 shows the initial capacities and cycle performance of the cells assembled from the Fe-containing positive plates. As shown in Figure 11a, when the content of Fe in the synthesized leady oxide was more than 223 mg kg⁻¹ of LO-0.02, the initial capacity of the cells exhibits a decreasing trend. Figure 11b shows the 100% DOD cycle performance of the cells. The curves indicate that the capacity of cell made from LO-0 without Fe impurity kept stable within 200 charge-discharge cycles and its discharge capacity dropped to 1.0 Ah at the 249^th charge-discharge cycles. However, the cell made from LO-0.005 exhibits a shorter charge-discharge cycle life, with its capacity decreased from 2.2 to 1.0 Ah after only 230 charge-discharge cycles. In addition, the cycle life of the cells was dramatically shortened with the further increase of Fe content, especially for the cell made from LO-0.05, LO-0.2, and LO-0.5. Figure 11c shows the charging and discharging curves of testing cells assembled from the positive plates made from LO-0 and LO-0.2 in the first charge-discharge cycle. The cell made from LO-0.2 shows faster voltage rise during charging and faster voltage drop during discharging than the cell made from LO-0, indicating that the Fe impurity has a negative effect on the charge storage property of positive plate.

Zoom In Zoom Out Reset image size

When the Fe element content in the leady oxide was more than 223 mg kg⁻¹ of LO-0.02, the microscopic morphology and phase composition of the PAMs changed significantly during curing, formation procedures, and charge-discharge cycles. As a result, the initial capacity of testing cell decreased significantly. Besides, the increased hydrogen and oxygen gassing rate caused by Fe impurity aggravated the PAM expansion and shedding during the charge-discharge cycles.^16,18 Therefore, the premature capacity loss of the testing cell was observed when the Fe element impurity content in the leady oxide was only 49 mg kg⁻¹ of LO-0.005.