1. Introduction
Zinc is an important metal resource with significant application value in many industrial areas [
1]. Compared to zinc oxide minerals, zinc sulfide minerals (sphalerite and marmatite) remain the primary source of zinc because the beneficiation technology used to extract zinc from them is economically and technically advantageous [
2,
3]. However, zinc sulfide minerals are commonly associated with pyrite in mineral deposits. Thus, the separation of sphalerite or marmatite from pyrite is a problem that must be faced during the beneficiation of zinc sulfide minerals [
4].
As the pyrite surface is hydrophilic in alkaline conditions [
5,
6,
7], separation is usually achieved by depressing pyrite in alkaline environment [
8,
9,
10,
11]. Zinc sulfide minerals have poor affinities towards xanthate and so they are usually activated using copper sulfate before being collected using xanthate. However, when the pyrite is depressed in alkaline conditions, the activation of sphalerite or marmatite using copper sulfate is also affected. This is because, in alkaline media, copious amounts of copper hydroxide are formed in the copper sulfate solution at the concentrations generally used (10
−6 to 10
−4 mol/L) [
12,
13,
14,
15,
16,
17].
The formation of copper hydroxide leads to two significant changes. First, the copper ion concentration is kept low, which means that the activation of zinc sulfide mineral through the ion-exchange process (as shown by Equation (1)) is weakened [
18,
19,
20,
21,
22,
23,
24]. Secondly, the activation effect caused by copper sulfideconverted fromadsorbed copper hydroxide (as shown by Equations (2) and (3)) becomes dominant [
19,
20,
21]:
The floatability of the zinc sulfide minerals after activation is determined by the ratio of surface hydrophobic species (such as copper sulfides, polysulfides, and elemental sulfur) and hydrophilic species (such as hydroxides) [
25]. As the process of transformation of copper hydroxide into copper sulfide is related to the dissolution of surface zinc hydroxide, the transformation process is relatively slow [
18]. Therefore, when the copper sulfate concentration is excessively high, the copper hydroxide adsorbed on the mineral surface cannot be completely transformed into copper sulfide within a limited activation time. As a result, the ratio of hydrophilic copper hydroxide on the mineral surface is increased and will result in the depression of the zinc sulfide minerals. Flotation experiments on zinc sulfide minerals (and related surface detection experiments) reported by Boulton [
14], Mirnezami [
26], and Fornasiero et al. [
20] all confirm that excessively high concentrations of copper sulfate under alkaline conditions cause depression of zinc sulfide minerals because of the adsorption of large amounts of copper hydroxide.
The problem of activating zinc sulfide minerals using copper sulfate underalkaline conditions stems from defects in the activation process related to copper hydroxide formation. The activation mechanism when using CuSO
4 in alkaline environments has some flaws: (1) the activation rate is affected by the dissolution of zinc hydroxide on the mineral surface; and (2) the activation effect is sensitive to the copper concentration [
18,
20].
Therefore, it makes sense to consider taking measures to mitigate the above-mentioned effects and strengthen the activation effect of copper sulfate on zinc sulfide minerals under alkaline conditions. Some research has already been done in this area. One method in particular, adding ammonium chloride, has been used to enhance the activation of copper sulfate towards marmatite under alkaline conditions [
27,
28]. However, these studies have merely verified the feasibility of the method—no attempts were made to investigate the mechanisms involved in great depth.Therefore, in this paper, research is carried out on the mechanism by which ammonium chloride strengthens the activation of marmatite using copper sulfate under alkaline conditions. Micro-flotation, adsorption test, X-ray photoelectron spectroscopy (XPS), and copper-concentration measurements are the tools used to accomplish this.
2. Experimental
2.1. Materials and Reagents
The marmatite sample used in this study was obtained from the Dulong Mine in Yunnan Province, China. The sample was dry-ground using a porcelain ball mill and then dry-screened to select particles of size 0.075 ± 0.044 mm for experimental use. The chemical composition of the sample is presented in
Table 1. The marmatite sample was stored in a glass bottle and protected under N
2. In all experiments, the marmatite samples were first scrubbed under ultrasonication for 5 min and then washed three times with deionized water to obtain fresh, consistent, mineral surfaces.
All the water used was purified using an ultra-pure water machine (UPT-I-10T, ULUPURE, Xi’an, China) with a (RO) reverse osmosis unit. The specific resistance of the Milli-Q water is 18.2 MΩ ·cm−1. The copper sulfate (CuSO4·5H2O), ammonium chloride (NH4Cl), sodium hydroxide (NaOH), and ethylenediaminetetraacetic acid disodium salt (disodium EDTA) used in the experiments were all of analytical purity. The collector, sodium isobutyl xanthate, was purified by dissolving it in acetone and then recrystallizing it from ether. A fresh sample was prepared in solution on the day of use to avoid oxidation and decomposition.
2.2. Flotation Experiments
Flotation experiments were conducted in a 50 mL Hallimond tube with 1 g of marmatite sample (0.075 ± 0.044 mm) per test. After being scrubbed under ultrasonication for 5 min and then washed three times with deionized water, the marmatite sample was first conditioned for 4 min in the tube with 40 mL of pH = 9 (adjusted by 0.001 mol/L NaOH) activator solution with desired concentration. It was then conditioned for another 2 min with addition of 4 mg/L of sodium isobutyl xanthate. After that, flotation was performed for 10 min with an air flow rate of 0.010 L/min. After flotation was finished, the floated and non-floated fractions were filtered, dried at 40 °C, and then weighed to calculate recovery. Each test was repeated three times and the average recovery rate calculated for final reporting. The experimental errors of recovery are quantified using standard deviation.
2.3. Adsorption Tests
The residual concentration method was used to measure the total amountof copperadsorbed. Extraction with disodium EDTA was used to achieve selective extraction and determine the amount of copper hydroxide adsorbed onto themarmatite surface.
A sample (1 g) of marmatite after washing wasfirst conditioned in 20 mL corresponding pH = 9 (adjusted by 0.001 mol/L NaOH) activator solution (3 × 10
−5 mol/L CuSO
4/cuprammonium solution (3 × 10
−5 mol/L CuSO
4 plus 6 × 10
−4 mol/L NH
4Cl)) for a required amount of time. After conditioning, the suspension was filtered and the filtrate was stored in a waterless and clean tubeas “Sample No. 1” for testing—thefilter residuewasquickly transferred to a 50 mL waterless and clean vial for extraction. The extraction was conducted in 20 mL 3 wt % disodium EDTA solution for 20 min. After the extraction was complete, the suspension was filtered and the filtrate was stored in a waterless and clean tubeas “Sample No. 2” for testing. The disodium EDTA solutionwas purged with nitrogen before use to reduce the oxygen content of the solution to prevent oxidation of the marmatite during the extraction process. Extraction was performed at pH = 9 (adjusted by 0.001 mol/L NaOH) to ensure a good extraction effect was obtained [
29].
The copper and zincconcentration was measured using inductively coupled plasma atomic emission spectrometry (ICP-AES) (ICPS-1000II, Shimadzu, Kyoto, Japan). The sample to be tested is acidified with hydrochloric acid before testing to prevent hydrolysis and precipitation of metal ions.The total amount of copperadsorbed and the extracted and unextracted copper on the marmatite surface can be calculated using the following expressions:
where
C0,
C1, and
C2 are the initial copper concentration, the copper concentration of Sample No. 1, and that of Sample No. 2, respectively. The volumes
V1 and
V2 are the volume of the activation solution used for treatment and the disodium EDTA solution used for extraction, respectively, and
M is the mass of the solid sample used in each adsorption test.
The extracted zinc can also be calculated using the Formula (5). However, in this case, C2 is the zinc concentration of Sample No. 2.
The experimental errors ofadsorption amount are quantified though standard deviation and are presented in the graph.
2.4. XPS Analysis
To compare the composition of the material produced on marmatite surface conditioned with copper sulfate solution and that conditioned with cuprammonium solution, XPS measurements were undertaken. The XPS measurements were carried out using a spectrometer (PHI5000 Versaprobe II, ULVAC-PHI, CHIGASAKI-SHI, Kanagawa, Japan) equipped with a monochromatic Al K X-ray source at a pass-energy of 46.95 eV. The pressure in the analyzer chamber was 5 × 10−9 mba during analysis. The detection process included a survey scan to identify the chemical components present and high-resolution scans to detect the levels of certain elements.
First of all, 1 g of washed marmatite was stirred for 10 min with the following solutions, respectively (each at pH = 9):
- (a)
Deionized water,
- (b)
6 × 10−4 mol/L NH4Cl,
- (c)
3 × 10−5 mol/L CuSO4,
- (d)
3 × 10−5 mol/L CuSO4 + 3 × 10−4 mol/L NH4Cl,
- (e)
3 × 10−5 mol/L CuSO4 + 6 × 10−4 mol/L NH4Cl.
After being conditioned, the suspension was filtered and then the filter residue was rinsed with deionized water three times. After flushing, these samples were dried in a vacuum drying oven at 30 °C. After drying, these samples were correspondingly labeled “(a)–(e)” and their XPS spectra were recorded.
2.5. Concentration of Copper in Solution
The 3 × 10−5 mol/L copper sulfate solution and cuprammonium solutions with different molar ratios of ammonium chloride to copper sulfate were prepared. Then, all solutions were adjusted to pH = 9 and centrifuged at 3000 rpm (using a TL4.7W centrifuge, Shanghai Shengkalan Industrial Co., Ltd., Shanghai, China) for 30 min. After that, the solutions were filtered by the syringe filter (the pore size of the filter is 0.22 µm) and the copper concentrations of filtrates were measured using ICP-AES.
3. Results
3.1. Flotation Results
At pH = 9, under the condition of 4 mg/L sodium isobutyl xanthate and a fixed M(NH
4Cl)/M(CuSO
4) of cuprammonium solution, the flotation experiments with and without addition of ammonium chloride were conducted and the results are shown in
Figure 1.
As shown in
Figure 1, when the copper sulfate concentration in the range of 2~6 × 10
−5 mol/L, the recovery is significantly higher when the copper sulfate has ammonium chloride added to it. When the copper sulfate concentration exceeds 6 × 10
−5 mol/L, the recovery rates with and without addition of NH
4Cl both show a downwards trend, in agreement with previous research [
14,
20,
26]. This is expected to be due to the marmatite surface being coated with copper hydroxide formed in the presence ofexcessive amounts of copper sulfate under mild alkaline conditions.
To further explore the effect of adding ammonium chloride on the activation of marmatite by copper sulfate, the flotation rate experiment under activation of 3 × 10
−5 mol/L CuSO
4 and 3 × 10
−5 mol/L CuSO
4 + 6 × 10
−4 mol/L NH
4Cl are conducted, the experimental results are shown in
Figure 2.
Figure 2 shows that the addition of ammonium chloride leads to an increase in flotation rate during the first six minutes of flotation. The flotation results clearly show that adding ammonium chloride improves the flotation rate of marmatite.
3.2. Adsorption Test Results
3.2.1. Copper Adsorption
The amount of copper adsorbed is a parameter that allows us to intuitively assess the activation effect of the mineral. According to previous research [
12,
19,
20], the copper adsorbed onto the surface of the sphalerite and marmatite when they are activated by copper sulfate under alkaline conditions usually appears in the form of both copper sulfides and copper hydroxide. Obviously, these different copper compounds will have a completely different effect on the hydrophobicity of the mineral surface. Therefore, the activation effect cannot be properly characterized by simply measuring the total amount of copper adsorbed onto the surface of the mineral.
Rumball et al. [
29] found that disodium EDTA can be used to selectively leach metal oxides, hydroxides, carbonates, and sulfates from sulfide minerals. Fornasiero and Gerson [
12,
20] used disodium EDTA to extract copper hydroxide from the surface of sphalerite in their research on how copper sulfate activates the sphalerite. The selective extraction effect of the disodium EDTA was verified in their research using XPS.
In order to separately determine the amounts of copper sulfide and copper hydroxide adsorbed onto the marmatite surface, selective extraction of the adsorbed copper hydroxide component was carried out using disodium EDTA. Under alkaline conditions (pH = 9), the amounts of copper sulfide and copper hydroxide adsorbed on the marmatite after conditioning in 3 × 10
−5 mol/L copper sulfate, with and without the addition of ammonium chloride, were measured and compared as a function of time (
Figure 3).
In
Figure 3, the unextracted copper corresponds to copper sulfide and the extracted copper corresponds to copper hydroxide. As the figure shows, the amount of copper hydroxide and copper sulfide adsorbed increase with time in both of the two solutions. For a given amount of conditioning time, it is clear that the addition of ammonium chloride leads to a significantly larger amount of copper sulfideadsorbed onto the marmatite compared to conditioning using just copper sulfate. As for the hydroxide component, the addition of ammonium chloride clearly suppresses the amount of copper hydroxide formed over time.
3.2.2. Zinc Hydroxide Adsorption
The method of EDTA extraction was also used to extract the surface adsorbed zinc hydroxide, and the results are presented in
Figure 4.
As the figure shows, compared with marmatite treated in deionized water, less zinc hydroxide adsorbed on the surface of marmatite conditioned in NH4Cl; Compared with marmatite treated in 3 × 10−5 mol/L CuSO4, zinc hydroxide adsorbed on the surface of marmatite conditioned in cuprammonium solution (3 × 10−5 mol/L CuSO4 + 6 × 10−4 mol/L NH4Cl) decreases with increasing ammonia chloride addition. It can be concluded from theseresults that adding NH4Cl can reducing zinc hydroxide adsorbed on marmatite surface.
3.3. XPS Studies
To explore the differences occurring in marmatite surfaces conditioned in copper sulfate, with and without the addition of ammonium chloride, marmatite samples conditioned in different solutions were subjected to XPS analysis.
The XPS spectraand surfaceatomic ratios of sample “(a)–(e)”were obtained and calculated based on the Multi-Pak Spectrum software (V 9.0). TheC1s spectral peak at 284.8 eV was used for calibration of all spectra.The XPS survey scan results are listed in
Table 2.
From the above table, it can be seen that the most obvious changes in surface content occurred on oxygen and copper and no significant change in the content of other elements was found.Whether in the case of deionized water treatment or copper sulfate treatment, the oxygen content on the mineral surface shows a significant drop after the addition of ammonium chloride. Thedrop in surface oxygen contentmay result from decreased content of surface-adsorbed hydroxide. In the case of conditioning in copper sulphate, the amount of copper adsorbed on the mineral surface increases with the increase in the amount of ammonium chloride added. The results obtained from the XPS survey scan resultsare consistent with the EDTA extraction results, which proves the reliability of XPS analysisto some extent.
3.3.1. Surface Nitrogen Analysis
Figure 5 shows the N 1s XPS spectra of samples “(c)–(e)”. It can be seen from the figure that the N 1s XPS spectra of the three samples are all similar and lack obvious peaks at band energies of 398. 60 eV (NH
3) [
30] and 401. 80 eV (NH
4+) [
31] or other positions. This means that none of these samples contained significant amounts of NH
3 and NH
4+ or other nitrogenous substances, such as Cu(NH
3)
n2+, on the surfaces of the particles contained within. This result excludes the possibility that the marmatite is activated through the adsorption of NH
3, NH
4+, or Cu(NH
3)
n2+.
3.3.2. Surface Zinc Analysis
Based on the Zn 2p
3/2 binding energies given in the literature [
32,
33,
34], the following three zinc species were identified on the surface of the samples: ZnO (1021.20 eV), ZnS (1022.00 eV), and Zn(OH)
2 (1022.70 eV). The ZnO component is probably formed by the dehydration of zinc hydroxide during the drying process [
35]. The results of peak fitting of the Zn 2p
3/2 data are shown in
Figure 6 and the zinc content is summarized in
Table 3.
Compared to sample (a), the ZnO and Zn(OH)
2 content of sample (b) is significantly lower (the total reduction amounts are 35.46%). Comparing the ZnO and Zn(OH)
2 content of samples (c)–(e), wherein increasing amounts of NH
4Cl are added to the copper sulfate, the amounts of ZnO and Zn(OH)
2 present are seen to decrease. The characterization and comparison of the surface zinc species given above shows that adding ammonium chloride does reduce the zinc hydroxide content of the marmatite surface. These results are consistent with the adsorption results given in
Section 3.2.2.
3.3.3. Surface Copper Analysis
According to the Cu 2p
3/2 binding energies given in the literature [
36,
37], the following three copper species can be identified on the surface of the samples: Cu
2S (932.20 eV), CuO (933.60 eV), and CuS (935.00 eV). The CuO may be formed through the dehydration of the adsorbed copper hydroxide during the vacuum-drying process [
38]. The Cu
2S is generated via the reduction of CuS, according to the results of previous research [
39]. The peak fitting results for the Cu 2p
3/2 XPS data are shown in
Figure 7 and the corresponding copper content is listed in
Table 4.
According to previous XPS research [
40], the Cu(II) peaks in the Cu 2p XPS spectra should show strong satellite peaks at 943.4 and 961.9 eV. As can be seen from
Figure 8, these satellite peaks are all weak and gradually decline in significance from sample (c) to sample (e). This indicates that there is only a small amount of Cu(II) on the surfaces of these samples, which is consistent with the abundances listed in
Table 4.
Comparing the copper content of the samples, it can be seen that the relative amount of CuO present gradually decreases as the amount of NH
4Cl added to the copper sulfate increases. The foregoing XPS analysis of the copper species on the mineral surfaces shows that adding NH
4Cl to the copper sulfate reduces the amount of copper hydroxide on the marmatite surface. These results are consistent with the adsorption results given in
Section 3.2.1.
3.4. Solution Components
The composition of the solution directly affects adsorption onto the mineral and products formed thereon. Thus, it is a crucial aspect of the flotation process. The differences in the composition of copper sulfate solution, with and without ammonium chloride, will therefore have a significant impact on the activation of the marmatite. Hence, it makes sense to explore the difference in compositions of the two solutions.
Under alkaline conditions (pH = 9), the copper concentration of a copper sulfate solution (3 × 10
−5 mol/L) with different amounts of NH
4Cl added were measured. The results are shown in
Table 5.
As can be seen from
Table 5, for a given pH and copper sulfate concentration, the copper concentration increases as the amount of ammonium chloride added increases. The concentrations listed in
Table 4 are all less than the copper concentration corresponding to 3 × 10
−5 mol/L (1.91 mg/L), which implies that all the solutions contain copper hydroxide precipitate.
In order to help our discussion of the effect of adding ammonium chloride to the copper sulfate solution, the compositions of s copper sulfate solutions with addition of different amounts of NH
4Cl were calculated using Visual MINTEQ, a chemical equilibrium model frequently used in related fields to calculate metal speciation, solubility equilibria, adsorption, etc. in aqueous solution [
41,
42,
43,
44,
45]. The latest version (3.1) was used in this work.
Under alkaline conditions (pH = 9), and with a copper sulfate concentration of 3 × 10
−5 mol/L, the results of the equilibrium calculations when different amounts of NH
4Cl are present are shown in
Figure 9. It can be seen from the figure that, as the amount of NH
4Cl increases, the proportion of dissolved copper increases. This is consistent with the results of our copper concentration measurements. More specifically, the activities of the Cu(NH
3)
n2+, CuCl
−, and CuCl
2(aq) components all increase while the activities of the other copper-containing components are almost unchanged. Considering the greater magnitudes of the activities of the Cu(NH
3)
n2+ species (compared to CuCl
− and CuCl
2(aq)), these ions are believed to be the ones responsible for the increase in the copper concentration as NH
4Cl is added.
The experimental results, combined with the solution chemistry calculations, show that, at pH = 9, the concentration of Cu2+ in both systems (3 × 10−5 mol/L CuSO4 and cuprammonium solution consisting of 3 × 10−5 mol/L CuSO4 + 6 × 10−4 mol/L NH4Cl) at equilibrium are determined by the solubility product Ksp of Cu(OH)2(s). The significant difference between these two systems is that part of the copper hydroxide is dissolved to form Cu(NH3)n2+ in the cuprammonium solution.
4. Discussion
The XPS analysis and EDTA extractionof the zinc species on the mineral surfaces has confirmed that adding NH4Cl to the copper sulfate can reduce the amount of zinc hydroxide formed on the marmatite surface. The theoretical explanation for this phenomenon is suggested to be: the NH3(aq) can promote the dissolution of the zinc hydroxide through a complexation reaction between NH3 and Zn2+.
The dissolution of Zn(OH)
2(s), with and without the involvement of NH
3(aq), can be represented by the equations:
Equation (7) can be obtained by combining Equations (8) and (9), and so the equilibria constants are such that:
The complexation reactions of Zn
2+ and NH
3 are listed in
Table 6, which shows that the constants corresponding to
K9 are large. Hence,
K7 ≫
K8. This comparison of equilibria constants theoretically justifies the enhancement effect of NH
3(aq) on the dissolution of zinc hydroxide.
According to previous studies [
19,
20,
21,
39], the activation of zinc sulfide minerals under alkaline conditions is mainly achieved through the following process:
The Zn(OH)2 formed on the marmatite surface during activation will undergo dissolution and dispersion, the extent of which affects the copper hydroxide replacement process. This then determines the degree of hydrophobicity of the marmatite surface.
The copper adsorption measurements in
Section 3.2 and copper XPS analysis of the surface species in
Section 3.3.3 both indicate that adding NH
4Cl to the copper sulfate solution leads to a reduction in the amount of copper hydroxide on the marmatite surface. In addition to the reactions replacing the copper hydroxide on the surface, the reduction in surface copper hydroxide could also be due to the promotion of zinc hydroxide dissolution, which accelerates the reactions replacing the copper hydroxide. In the copper adsorption measurements, the amount of copper hydroxide increased more slowly (and copper sulfide increased more rapidly) on the surface of marmatite conditioned using copper sulfate with added ammonium chloride, which coincides with this explanation.
In accordance with the above discussion, the following mechanism can be hypothesized. Adding ammonium chloride to the copper sulfate solution promotes the dissolution of zinc hydroxide on the marmatite surface and then facilitates the transformation of adsorbed copper hydroxide into a stable activation product (copper sulfide). This causes an increase in the amount of surface copper sulfide, which can easily adsorb thiol collectors and decreases the amount of hydrophilic zinc hydroxide and copper hydroxide on the surface. The combination of these two effects eventually results in a significant increase in the recovery of marmatite during flotation.
Further analysis of the copper adsorption results reveals a noteworthy phenomenon: The amount of copper sulfide present increases much faster than the rate at which the amount of copper hydroxide decreases when conditioned with copper sulfate and ammonium chloride (compared with copper sulfate on its own).
The above phenomenon and the observed differences in the solutions’ components suggests that there may be another mechanism occurring simultaneously: the Cu(NH3)n2+ formed after adding ammonium chloride has the effect of storing and releasing Cu2+—hence, it can “buffer” the tendency of the copper ion concentration to drop and then promote the Cu2+ adsorption process.
Although precipitated copper hydroxide is the main form of copper under alkaline conditions, a certain amount of the Cu
2+ remains in solution which is still able to activate the marmatite through the copper-zinc ion-exchange mechanism [
18,
19,
20,
21].
Regardless of which solution is used (copper sulfate or cuprammonium solution), during the absorption of copper ions onto the marmatite surface, there is continuous shifting of the dynamic equilibria between the components of the solution. In the cuprammonium solution, the Cu
2+ equilibrium processes occurring during the adsorption process can be represented by the simplified scheme shown in
Figure 10.
Several aspects of the processes illustrated in
Figure 10 are in need of further explanation: (1) in the diagram,
C2 <
Csaturated. The reason for this is that the copper hydroxide has not yet decomposed to release and replenish the Cu
2+ concentration and so the total copper concentration is reduced due to the adsorption of copper ions onto the mineral surface; (2) the Cu(NH
3)
n2+ ions are uniformly dispersed in the liquid phase, while the copper hydroxide exists as a precipitate or colloid in the solid state [
27]. Thus, the contact area between the former and the marmatite surface is much greater than that of the latter. When copper ions are adsorbed onto the mineral surface, the decrease in copper ion concentration will initially affect the dissociation equilibria of the Cu(NH
3)
n2+ species in the liquid phase. The dissolution equilibrium of the copper hydroxide will be affected until the fall in copper concentration spreads to the liquid phase near the solid surface of the copper hydroxide. Therefore, when the concentration of Cu
2+ decreases, Cu(NH
3)
n2+ will decompose before the copper hydroxide decomposes.
As can be seen from the processes shown in
Figure 10, Cu(NH
3)
n2+ has the function of storing and releasing Cu
2+. It can maintain the concentration of Cu
2+ and thereby accelerates the adsorption of Cu
2+.