Study on Selective Removal of Impurity Iron from Leached Copper-Bearing Solution Using a Chelating Resin

Abstract

In order to selectively remove iron from copper laden solution after leaching but prior to electrowinning, equilibrium, kinetic, and thermodynamic studies have been conducted on an a chelating resin of Rexp-501 at pH 1.0 and at various temperatures. Both Langmuir and Freundlich models were investigated, with the Langmuir model proving to be more suitable for fitting iron removal performance, with little influence from copper concentration. Compared with the pseudo first order kinetic model, the pseudo second order kinetic model fitted the dynamic adsorption process better, indicating a chemisorption mechanism. Fourier transform infrared spectroscopy (FT-IR) results indicated that C=O from carbonyl group played a key role in combining with iron and can be regenerated and reused. However, the C=O of the acylamino group combining with iron was not able to be released after oxalic acid was applied.

1. Introduction

The hydrometallurgical processing of copper-bearing minerals has attracted much attention due to its relatively cost-effective and environmentally friendly properties as compared with its pyrometallurgical counterpart [1,2,3,4,5]. Due to the complexity of the naturally formed minerals, leaching of other gangue-containing impurity elements or minerals normally accompanies the dissolution of copper minerals, resulting in the significant release of some cations, especially iron that dissolved remarkably with copper from the copper-iron minerals such as chalcopyrite. As chalcopyrite accounts for approximately 70% of the copper reserve worldwide [6,7], the separation of these cations from copper pregnant solutions is a key step for successful copper enrichment. Most studies have tried to recover the valuable copper using solvent extraction [8,9,10,11]. However, less attention has been paid to removing the impurity cations, especially the great amount of iron present in the laden solution, thereby leaving a much purer copper solution. As the chalcopyrite leaching solution containing impure iron would lead to a problem for the subsequent extraction of copper in the electrowinning process, the separation of iron from copper-bearing solutions obtained in the hydrometallurgical process of copper-containing minerals such as chalcopyrite [2,5] is of great concern worldwide.

The simplest way to remove iron is to add chemicals to increase the solution pH to precipitate iron. However, the precipitated solution with high pH values should be treated with acid again prior to reusing for further chalcopyrite leaching. Therefore, the techniques having the capability to remove iron but not increasing solution pH are desired in the hydrometallurgical processing of copper minerals, especially at low pH conditions that are suitable for acid leaching. Considering the technologies applied in, e.g., wastewater treatment, many adsorbents used in heavy metal removal from wastewater, such as activated carbon [12], graphite oxide [13], ion-exchange [14], and membrane technologies [15], have been applied to treat heavy metals in wastewater, although some industrial byproducts have also been utilized for heavy metal treatments [16,17,18].

Recently, the selective adsorption strategy using chelating resins has been paid enormous attention due to its simplicity, high efficiency, and low cost during the adsorption processes [19,20,21]. Polymerization of conventional chelating monomers such as acrylic acid, methacrylic acid, and vinylpyridine has been extensively investigated to develop suitable resins to selectively adsorb metal ions [22,23,24,25]. This study was therefore aimed to investigate the selective removal capability of iron by a chelating resin at a low pH condition. In addition, the regeneration capacity of the resin was also investigated.

2. Materials and Methods

2.1. Chemicals

The chelating resin Rexp-501 was purchased from Ruilekang Separation Technology, Beijing, China. The size of the white spherical resin ranges from 178 to 840 μm. The inorganic and organic chemicals including CuSO₄, Fe₂(SO₄)₃, H₂SO₄, HCl, oxalic acid (C₂H₂O₄), and citric acid (C₆H₈O₇) were of analytical grade and purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. In addition, KBr used for Fourier transform infrared spectroscopy (FT-IR) analysis was in spectral purity. All chemicals were used as received without any modification.

2.2. Adsorption Experiment

All adsorption experiments were carried out using fresh solutions prepared from analytical grade reagents and MilliQ water with a resistivity greater than 18.2 MΩ·cm. Iron loading solutions of various concentrations (170–10,000 mg·L⁻¹) were prepared using ferric sulfate (Fe₂(SO₄)₃), with pH being adjusted to 1.0 using H₂SO₄. For the iron adsorption experiment in the presence of copper, a mixture containing both copper using CuSO₄ and iron using Fe₂(SO₄)₃ was synthesized.

For the adsorption experiment, 1–50 g Rexp-501 was added into 100 mL solutions with various iron and copper concentrations and shaken in an air bath oscillator at 298–328 K and 150 rpm. In addition, the kinetic studies were sampled at a reasonable interval (i.e., 0.5, 1, 2, 3, 4, 5, 8, 10, 15, 20, 30, 60, and 90 min). After adsorption, a 1 mL solution was taken out for a solution assay.

The amount of iron or copper adsorbed by Rexp-501 was determined according to Equation (1):

$$Q_{\mathrm{e}}=\frac{C_0{V}_0-C_{\mathrm{e}}{V}_{\mathrm{e}}}{W},$$

(1)

where Q_e is the amount of iron or copper adsorbed per unit weight of Rexp-501, C₀ and C_e are the iron concentrations before and after adsorption, and V₀ and V_e are the volumes of the solution before and after adsorption, respectively. W is the weight of the chelating resin added to the solution.

The regeneration experiments were conducted using 10 wt % various acids (Section 2.1 and Section 3.6, i.e., H₂SO₄, HCl, C₂H₂O₄, and C₆H₈O₇) to remove ions from saturated resin. After being generated in such a way, the resin was further used to test its adsorption capacity.

2.3. Measurements

The element concentrations were analyzed using a flame atomic absorption spectroscopy Agilent AA240FS (Agilent Technologies Inc., Santa Clara, CA, USA), while the functional groups of this resin were analyzed via Fourier transform infrared spectroscopy (FT-IR, Nicolet IS10, Thermo Fisher, Waltham, MA, USA). All the FT-IR samples were prepared mixed with spectral purity KBr (150 mg) and resin (1 mg), and pressed to form pellets prior to analysis.

3. Results and Discussions

3.1. Separation Efficiency

The synthetic solution with both iron and copper added at the same time were treated by a chelating resin named Rexp-501 to evaluate its separation efficiency of iron from copper solutions. Figure 1 shows the adsorption amounts of iron and copper under various concentrations, but under a constant temperature of 298 K. It is observed that the unit adsorption capacity for iron decreased gradually from 3.53 to 1.86 mg/g when the resin amount increased from 1 to 6 g/100 mL, but that for copper remained lower than 0.13 mg·L⁻¹ when iron and copper concentrations were both 170 mg·L⁻¹. When iron and copper concentrations increased from 170 to 680 mg·L⁻¹, the adsorption capacity of iron increased significantly, but that of copper did not follow the same trend, indicating that better separation efficiency of iron from copper-containing solutions can be achieved at greater iron and copper concentrations. This further verified that resin Rexp-501 is capable of separating iron from copper solutions at high concentrations.

Furthermore, Figure 2 shows the separation efficiency (iron and copper) of iron from copper solutions using the same concentration as shown in Figure 1. As indicated, the separation efficiency decreased gradually when the resin amount increased. However, the ratio between the adsorption amount of iron and copper increased when iron and copper concentrations increased from 170 to 680 mg·L⁻¹, indicating that greater iron and copper concentrations resulted in better iron and copper separation. This is very important in separating iron from chalcopyrite leaching solution in industry, as both copper and iron concentration were very high prior to the separation or extraction step. The above results indicated that the adsorption of iron was not influenced by the copper concentration, as the adsorption of copper was limited to a very low level. Therefore, further studies will only focus on the adsorption of iron in the absence of copper.

3.2. Effects of Resin Amount

Figure 3 summarizes the removal efficiency of iron from synthetic solution with different iron concentrations (170–10,000 mg·L⁻¹) using Rexp-501 at 298 K and 150 rpm for 90 min. As indicated, the iron removal percentage increased with the resin amount increased. Specifically, when the resin dosage was 10 g/100 mL, there was a significantly different iron adsorption under various iron concentrations, e.g., over 95% of iron was adsorbed when iron concentration was lower than 680 mg·L⁻¹. However, when iron concentration increased to 1700 mg·L⁻¹, only approximately 80% of the iron was removed, with the removal efficiency being further reduced when iron concentration increased, e.g., achieving less than 30% of iron removal when iron concentration was 10,000 mg·L⁻¹. It should be noted that the final removal of iron was almost 100% for all conditions when a sufficient amount of resin was added.

Furthermore, Figure 4 shows that the iron adsorption capacities of Rexp-501 decreased linearly as a function of resin dosage at any iron concentration. However, iron removal capacity increased when iron concentration increased, e.g., from approximately 3 mg·g⁻¹ at 170 mg·L⁻¹ to more than 30 mg·g⁻¹ at 10,000 mg·L⁻¹, indicating that sufficient utilization of resin and optimized iron removal rate were available at greater iron concentrations.

3.3. Adsorption Isotherms

Adsorption isotherms reveal the relationship between the equilibrium adsorption quantity and the equilibrium concentration (C_e) under a certain temperature. Both Langmuir (Equation (2)) and Freundlich (Equation (3)) equations were applied to fit the isothermal data:

$$\frac{C_{\mathrm{e}}}{Q_{\mathrm{e}}}=\frac{C_{\mathrm{e}}}{Q_{\mathrm{m}}}+\frac{1}{b{Q}_{\mathrm{m}}},\mathrm{and}$$

(2)

$$\log Q_{\mathrm{e}}=\log K_{\mathrm{f}}+\frac{1}{n}\log C_{\mathrm{e}},$$

(3)

where Q_e is the amount of adsorption per unit mass of sorbent, C_e is the equilibrium concentration of iron in the solution, Q_m is the amount of adsorbate at complete monolayer coverage, and b is the Langmuir isotherm constant related to the energy of adsorption. K_f and 1/n are Freundlich constants related to the adsorption capacity and intensity, respectively.

The Langmuir and Freundlich isothermal fitting results are shown in Figure 5 and

Table 1. Fitting results of Langmuir and Freundlich isotherms (R² is the correlation of determination).

Concentration (mg·L−1)	Langmuir Isotherm				Freundlich Isotherms
	Qm (exp) (mg·g−1)	Qm (cal) (mg·g−1)	b (L·mg−1)	R2	Kf	1/n	R2
170	3.1273	3.0562	0.1288	0.9655	1.3152	0.1598	0.8169
340	5.2458	5.2604	0.1020	0.9946	2.3496	0.1409	0.8369
680	9.5836	9.0580	0.0401	0.9806	4.0986	0.1167	0.9102
1700	14.5470	14.6628	0.0132	0.9861	7.8864	0.0749	0.8714
3500	21.5379	23.1481	0.0098	0.9838	4.2565	0.1969	0.9447
5000	26.1574	26.5957	0.0077	0.9977	12.5707	0.0831	0.9316
10,000	30.6689	31.2500	0.0021	0.9919	12.4825	0.0930	0.8924

. It is evident that the adsorption capacity of this resin fitted the experimental results well at any concentration investigated using the Langmuir isotherm model, e.g., increasing from approximately 3 mg·g⁻¹ at 170 mg·L⁻¹ to more than 30 mg·g⁻¹ at 10,000 mg·L⁻¹. In contrast, the isotherm constant b decreased from 0.1288 to 0.0021 when iron concentration increased from 170 to 10,000 mg·L⁻¹.

As 1/n < 1 indicates favorable adsorption while 1/n > 1 suggests cooperative adsorption, these 1/n values shown in Table 1 revealed that iron adsorption by Rexp-501 was favorable [26]. The K_f values varied but had a generally increasing trend with increasing iron concentration, indicating that greater adsorption capacity may be achievable at higher iron concentrations. However, as compared with the R² values obtained from Langmuir model fitting (greater than 0.96, Table 1), all the R² values calculated from the Freundlich model was lower, suggesting that the Langmuir model was more appropriate than the Freundlich model, and iron was adsorbed in a monolayer coverage status on the resin.

3.4. Thermodynamic Analysis

As shown in Figure 3, the addition of 25 g of resin can result in a 90% iron removal when iron concentration was 5000 mg·L⁻¹, further thermodynamic analysis was conducted to investigate the temperature effect. As shown in Figure 6, the iron adsorption process had three stages. Within the initial 6 min, a linear and rapid increase in iron removal by adsorption was observed, followed by a gradual increase until 30 min. The adsorption tended to be equivalent after 60 min, indicating a rapid adsorption resin [27]. It should be noted that the iron adsorption by resin varied insignificantly when temperature increased from 298 to 328 K within 6 min, but over 60% of iron was adsorbed within this period. In the following adsorption stage, the iron adsorption rate increased gradually with the increased temperature, although the final removal of iron (approximately 95%) was almost the same at 90 min, achieving a Q_e ranging from 18.43 to 18.74 mg·g⁻¹ from 298 to 328 K.

3.5. Adsorption Kinetics

To further analyze the adsorption kinetics, the Lagergren pseudo first order [28] and second order [29,30] equations (Equations (4) and (5), respectively) were applied in the modeling iron adsorption on this resin.

$$\log \left(Q_{\mathrm{e}}-Q_{\mathrm{t}}\right)=\log Q_{\mathrm{e}}-\frac{k_{1}t}{2.303};$$

(4)

$$\frac{t}{Q_{\mathrm{t}}}=\frac{1}{k_2{Q}_{\mathrm{e}}^2}+\frac{t}{Q_{\mathrm{e}}};$$

(5)

here, Q_e (mg·g⁻¹) is the equilibrium concentration of iron on resin, and Q_t (mg·g⁻¹) is the amount of iron adsorbed on resin at time t. k₁ (min⁻¹) and k₂ (g·mg⁻¹·min⁻¹) are the rate constants for the pseudo first order and pseudo second order equations, respectively. It should be noted that only the data collected within 30 min were used to fit the model, as there were no significant improvements in the adsorption at 60 or 90 min. Figure 7 shows the fitting of data shown in Figure 6, but restricted to the fast adsorption stage, i.e., the first 30 min, within which nearly 100% iron removal was achieved at 328 K.

As shown in

Table 2. Lagergren pseudo first and second order kinetics parameters for iron adsorption by resin Rexp-501.

T (K)	First Order				Second Order
	Qe (exp, mg·g−1)	k1 (min−1)	Qe (mg·g−1)	R2	k2 (g·min−1·mg−1)	Qe (mg·g−1)	t1/2 (min)	R2
298	18.43	0.0888	10.3412	0.9342	0.0253	18.5908	2.1257	0.9987
308	18.52	0.1020	10.4323	0.9683	0.0263	18.9322	2.0112	0.9983
318	18.65	0.1336	10.8665	0.9815	0.0267	19.5389	1.9134	0.9987
328	18.74	0.1557	11.4583	0.9926	0.0281	19.7472	1.8021	0.9987

, the equilibrium adsorption capacities using the Lagergren pseudo first order model differs from that observed experimentally at all temperatures tested (Figure 6 and Table 2), indicating that this model is not suitable for modeling the adsorption process. However, the Q_e calculated based on the Lagergren pseudo second order are very close to that derived from the experiment. In addition, all the R² values are greater than 0.998, indicating that the adsorption of iron by Rexp-501 follows the Lagergren pseudo second order kinetic model. This further confirms that iron adsorption by resin Rexp-501 is due to chemisorption [31].

The half-adsorption time, t_1/2, as the time for half of the adsorbate adsorbed on the adsorbent, is often used as a measure of the adsorption rate (Equation (6)):

$$t_{1/2}=\frac{1}{k_2{Q}_{\mathrm{e}}}.$$

(6)

The t_1/2 values were present in Table 2. With the increase in temperature from 298 to 328 K, the half-adsorption time decreased gradually, from 2.1257 to 1.8021 min. The fast adsorption rate in the initial stage was probably due to the availability of a large number of chelating sites.

As the pseudo second order model fitted the adsorption process well, the rate constant k₂ was used to calculate the adsorption energy of activation using the Arrhenius equation (Equation (7)):

$$\ln k_2=-\frac{E_a}{2.303RT}+\mathrm{constant},$$

(7)

where k₂ is the rate constant (g), R, E_a, and T are gas constant (8.314 J·mol⁻¹·K⁻¹), energy of activation (kJ·mol⁻¹), and the temperature in Kelvin, respectively. E_a can be calculated via the plot of lnk₂ against 1000/T (Figure 8). The activation energy for iron adsorption was found to be 6.16 ± 0.93 kJ·mol⁻¹, indicating that the adsorption was determined by a diffusion controlled process rather than a chemical reaction mechanism [6,32].

3.6. Regeneration and Reuse

In order to investigate the regeneration capacity of Rexp-501, the saturated resins were further desorbed using four acids, i.e., H₂SO₄, HCl, oxalic acid, and citric acid. Figure 9 showed that more than 60% iron was removed by adding 10 wt % acids given above, except for citric acid. It is evident that the highest removal efficiency was achieved when oxalic acid was added, following by HCl and then H₂SO₄, with citric acid being the lixivant with the lowest iron removal efficiency (less than 40%).

Further desorption kinetic studies by oxalic acid (Figure 10) demonstrated that iron can be removed very quickly, e.g., approximately 90% of the iron was removed within the initial 5 min. However, in the following 15 min, iron removal increased gradually, achieving a 94.6% removal at the end of 20 min, which was further increased to 99.8% when an additional 10 mL oxalic acid solution was applied after filtration (data not shown in Figure 10). The greater iron removal was achieved when the second desorption solution was used, probably due to the equilibrium of iron in the first desorption solution.

The regenerated resin was further used to adsorb iron solution with a concentration of 5000 mg·L⁻¹ to study its reusability. The results showed that the unit capacity of the regenerated Rexp-501 was 14.63 mg·g⁻¹, lower than 19.58 mg·g⁻¹ of the fresh resin sample, indicating that a portion of the metal adsorption capacity of Rexp-501 was lost, probably due to the reduced amount of functional groups on the resin surface that cannot be released from bonded iron via acid treatment. It should be noted that the capacity of 14.63 mg·g⁻¹ lasted for at least 10 times in the manner of use and regeneration using oxalic acid, indicating a good capacity of regeneration.

3.7. Adsorption Mechanisms

The adsorption mechanism can be determined through themodynamic parameters such as enthalpy (∆H), entropy (∆S), and free energy (∆G), according to Equations (8)–(11):

$$K_c=\frac{C_{Ae}}{C_e},$$

(8)

$$\Delta G=-RT\ln K_c,$$

(9)

$$\ln K_c=\frac{\Delta S}{R}-\frac{\Delta H}{RT},\mathrm{and}$$

(10)

$$\Delta G=\Delta H-T\Delta S$$

(11)

where $K_c$ is the thermodynamic equilibrium constant, and $C_{Ae}$ and $C_e$ are the adsorbed and equilibrium concentrations of iron in the solution. ∆G is the free energy of adsorption (J·mol⁻¹), T is the temperature in Kelvin, R is the universal gas constant, which is 8.314 J·mol⁻¹·K⁻¹, ∆H is the heat of adsorption, and ∆S is the standard entropy change. The ∆H (8122.78 J·mol⁻¹) and ∆S (49.31 J·mol⁻¹·K⁻¹) values can be obtained from the slope and intercept of a plot of lnK_c against 1/T. The thermodynamic parameters are shown in

Table 3. The thermodynamic parameters.

T (K)	$\mathbf{\Delta }\mathit{G}$(exp) (J·mol−1)	$\mathbf{\Delta }\mathit{G}$(cal) (J·mol−1)	${\mathit{K}}_{\mathit{c}}$
298	−6613.66	−6588.58	14.4121
308	−7036.82	−7081.67	15.5904
318	−7582.72	−7574.75	17.5790
328	−8078.50	−8067.84	19.3186

.

The negative ∆G value indicated that the adsorption of iron was a spontaneously occurring process at all temperatures. The value of ∆G were more negative when temperatures increased, indicating greater efficient adsorption at greater temperature. The positive ∆H and increased K_c values with the increased temperatures indicated that the adsorption of iron was an endothermic process [32].

Figure 11 demonstrates the FT-IR spectra of the fresh and iron adsorbed resins at 298 and 328 K as well as the resin desorbed using oxalic acid for 20 min at 298 K. The adsorption at 1733 cm⁻¹ in Figure 11a indicates the presence of ester carbonyl groups (C=O) due to stretching vibration [33,34]. A strong peak at 1655 cm⁻¹ was ascribed to the stretching vibrations of C=O in acylamino groups [33], while the peak at 1406 cm⁻¹ was ascribed to C–N stretching vibration [35]. In addition, the peak at 1451 cm⁻¹ suggested the deformation vibrations of metal hydroxyl groups and interstitial water (δ₃ (M–OH) + H₂O inters) [36].

When iron was adsorbed at 298 and 328 K (Figure 11b,c, respectively), the peaks of ester carbonyl groups shifted to 1731 and 1726 cm⁻¹, respectively, indicating the adsorption of iron with the –COO⁻ group, which was further confirmed by the desorbed resin shown in Figure 11d where a peak at 1733 cm⁻¹ is present again. The further shift observed at 328 K indicated that more iron was adsorbed by the carbonyl functional groups. The presence of 1733 cm⁻¹ observed in the desorbed resin (Figure 11d), which can be further used to adsorb metal gain (Section 3.6), indicated that the stretching vibration of the ester carbonyl groups played an important role in iron adsorption and can be regenerated after acid leach. However, the functional groups at 1655 cm⁻¹ disappeared after adsorption (Figure 11b,c, respectively) and regeneration (Figure 11d, respectively), suggesting that the stretching vibrations of the H₂N=C=O group cannot be regenerated after oxalic acid leach. Therefore, the functional groups can be reused in the iron adsorption due to the presence of ester carbonyl groups (C=O) on the Rexp-501 surface.

4. Conclusions

The present study shows that the chelating resin Rexp-501 is an effective adsorbent for the selective removal of iron from copper-containing solutions. Iron adsorption follows both Langmuir isotherm and pseudo second order kinetic models, indicating that iron adsorption by Rexp-501 follows a chemisorption mechanism. In addition, this resin adsorbs iron in a very fast manner. The adsorption capacity of iron via Rexp-501 increases with increased temperature, and further thermodynamic analysis indicates a spontaneous and endothermic adsorption process. Most of the iron adsorbed on the saturated resin is successfully removed by various acids, with oxalic acid being the most effective agent. Moreover, after 10-fold regeneration and reuse, the resin still presents good capacity for iron removal, with a portion of the iron removal capacity being lost. FT-IR analysis indicates that this portion is probably due to the H₂N=C=O group that cannot be released after combining with iron. In contrast, the ester carbonyl groups (C=O) have a more important role in removing iron and can be regenerated and reused. Further studies can be focused on the quantity identification of these functional groups.