The concentrate was washed to remove residual reagents carried over from the flotation stage (e.g., surfactants, depressants, frothers, etc.). 1 L of distilled water was used for every 100 g of concentrate. First, water and concentrate were mixed in a reagent flask and stirred at 60 °C for 30 min. After this, the mixture was filtered and then the concentrate was dried at 45 °C for 48 h.

Once the concentrate was washed and dried, it was separated into four particle sizes, using a Ro-Tap and four ASTM sieve sizes: #100, #200, #270 and #400. The particle size ranges obtained are shown in Table 2.

Table 2

Variable	Value
Particle size (µm)	72; -150+75; -75+53; -53+38
HCl added (%)	5; 10 & 15

Source: The Authors

The pre-treatment consisted of adding three different doses of technical hydrochloric acid (37.0% pure) to 3 g samples of washed concentrate previously moisturized with distilled water in Petri dishes. These dosages corresponded to 5, 10 and 15% of the stoichiometric consumption of HCl by the bornite, considering the complete dissolution of bornite by the hydrochloric acid under the presence of oxygen, as showed in eq. (5). The amount of water added was calculated to reach an overall humidity of 20%.

After this, the samples were stored in an oven at 25 °C for 15 d, adding distilled water periodically to compensate gradual evaporation.

Every sample was doubly prepared to determine its mineralogical composition with powder XRD (X-Ray Diffraction) techniques once the pre-treatment was done. In other words, 24 samples (four particle sizes and three acid dosages, with their respective duplicates) were pre-treated at 25 °C for 15 d, where half were leached, and the other half were ground and stored for a posterior analysis by powder XRD. Also, samples of concentrate without pre-treatment for each particle size were leached and the leaching residues were analyzed by powder XRD and SEM (Scanning Electron Microscopy). In sum, a total of 28 samples were used in this research.

Once the pre-treatment was completed, each sample was orbital leached at 180 rpm, in a 500 mL closed cylindrical reactor for 24 h at 25 °C. 300 mL of leaching solution ([Cl^-] = 60 g·L^-1, pH = 1.5) were used per sample, using NaCl and H₂SO₄. Furthermore, control samples of concentrate with no pre-treatment for each particle size studied were leached under similar conditions. The oxidation-reduction potential (ORP) and the pH of the solution were measured with a pH/ORP meter (Hanna portable pH/ORP meter, model HI 2002) with a silver/silver chloride reference electrode while the leaching was carried out. The pH was adjusted during the process, adding sulfuric acid when it was necessary. Aliquots of 3 mL were taken at 1, 2 and 24 h, where Fe²⁺ and copper concentrations were measured by AAS.

The solutions were filtered when the leaching stage was completed, and the liquid phases were separated from the solid residues. These residues were then washed with distilled water and dried for 48 h. After this, they were analyzed with XRD techniques to determine their mineralogical composition.

Using SEM and XRD techniques, the mineralogical composition of T₈₀ = 72 µm and between -53+38 µm pre-treated with 5% and 15% HCl were determined.

Table 5 provides information about the mineralogical species obtained at the end of the 15 d of pre-treatment. Each sample contained chalcopyrite and soluble atacamite type compounds. Here, through eq. (6)-(10), a possible mechanism is proposed, in accordance with Dutrizac et al. [15]. Eq. (6) suggests that the first step in the ferric ion leaching of bornite is to transform it into non-stoichiometric bornite and copper ions, where “x” is believed to take values between 0 and 2. Samples pre-treated with 5% HCl showed presence of bornite after 15 d of pre-treatment, while samples pre-treated with 15% HCl did not. This can be explained by looking at eq. (7), This reaction was not completed for samples pre-treated with 5% HCl. In comparison, samples pre-treated with 15% HCl were fully transformed into chalcopyrite and nonstoichiometric chalcocite, the latter further oxidized through eq. (8), probably reaching an idaite composition (CuFeS₂•2CuS). This idaite stage probably decomposed into chalcopyrite and covellite, as proposed in eq. (9), which could explain the presence of Cu₃₄S₃₂. At some point, chalcopyrite might be leached as proposed in eq. (10), which would explain the detection of Cu₄Fe₅S₈ (Cu_0.8FeS_1.6) by the XRD. The values of x, y and z in this equation are not well clarified, but clearly x & y cannot be more significant than one and z bigger than two [41, 42].

Table 5

Source: The Authors

Since water in the mixture evaporated as the days passed, it had to be refilled constantly during pre-treatment, which could have produced pH variations, leading to an atacamite formation (see Fig. 1), as proposed in eq. (11)-(12).

The presence of nantokite (CuCl) can be related to possible potential variations of the mixture during the pre-treatment resting time, where if the potential goes under 0.6 V versus SHE, the Pourbaix diagram presented in Fig. 1 predicts its formation_. However, as the potential was between 0.4-0.7 V versus SHE during leaching, this copper salt does not represent a problem because it can be leached as shown in Figure 2 and eq. (13)-(15).

As displayed in Fig. 5, the observed tendencies regarding the copper ion dissolution, for all four particle sizes analyzed, were similar. Increasing the concentration of HCl in an aqueous solution during pre-treatment led to an improvement in dissolution after 24 h of leaching. It is important to highlight that the copper ion recovery from the samples without pre-treatment were the lowest among all the samples.

Figure 5 Source: The Authors

Table 6 shows the mineralogical composition of leaching residues. The samples pre-treated with less concentrated HCl solution contained non-stoichiometric bornite. This means that these samples did not have sufficient acid during the pre-treatment to complete the transformations to soluble species, which led to a less efficient copper dissolution during the posterior leaching stage. The main reaction for samples pre-treated with 5% HCl is represented by eq. (6) with “x” taking values probably between 0 and 2.

Table 6

Source: The Authors

On the other hand, bornite of samples pre-treated with 15% HCl was completely transformed into idaite during pre-treatment, therefore, in the leaching stage, sulfide present in idaite could have been oxidized by Fe(III) and Cu(II), which both get reduced, as proposed in eq. (16)-(21).

The atacamite formed during pre-treatment for all HCl concentrations used must have been almost completely transformed into copper ions and water, as proposed in eq. (22)-(24), which could explain why it did not appear in Table 6.

Covellite is a well-known refractory ore, and it cannot be oxidized by Fe(III). When covellite is in the medium (due to idaite decomposition), it is slowly leached by oxygen in solution as presented in eq. (25)-(27). Thus, a longer leaching time would be necessary to leach all the covellite.

Finally, as proposed in eq (10), part of the chalcopyrite might have been leached, more likely on samples pre-treated with higher HCl.

Comparing Fig. 6 with Fig. 7 below, one can see that the first one shows no visible effects of leaching on the particles, while in the latter there are a lot of smooth cracks on the surface of the particles. These fissures are deeper in the sample pre-treated with 15% HCl solution, which here corresponds to Fig. 8, where some micro pores surrounding the fractures are clearly visible.

Figure 6 Source: The Authors

Figure 7 Source: The Authors

Figure 8 Source: The Authors

Putting these results together with the leaching results from Fig. 5, it is feasible to relate the presence of fissures and pores to the copper dissolution. In fact, one can assume that these observations are due to the pre-treatment stage effect of the HCl, as for the sample without pre-treatment there was no evidence of cracks or micro pores detected by electron microscopy, hence making this stage of treatment crucial for the copper recovery. Thus, it can be said that the direct contact of an HCl solution with bornite particles is responsible for the fractures and pores due to the HCl highly corrosive properties. Carneiro & Leão [36] stated in their paper that Cl^- ions play a crucial role in the changes observed on the surface form and properties of the reaction product during the acid leaching of chalcopyrite, where “…in the presence of NaCl, surface area and porosity of the leach residue are higher than those obtained without sodium chloride”.

The formation of fissures and pores is beneficial for copper dissolution, increasing the internal surface of particles, allowing for deeper penetration of HCl solution within each particle and avoiding a further passivation of ore metals by forming air-oxidation stable complexes at the same time.

Considering the results presented in Table 5, in view of the particle size, it is evident that the composition between samples with the same particle size are almost identical.

Fig. 9 illustrates the copper dissolution vs. HCl dosage during pre-treatment stage, organized by particle size. For each dosage of HCl, the copper recovery was similar. This means that the particle size does not have an important role on the formation of soluble species, at least in the size range evaluated in this paper. As discussed earlier, liquid HCl during pre-treatment led to the presence of fissures and micro- the difference in particle size negligible for the size range studied. These results corroborate the results presented in Table 6.

Figure 9 Source: The Authors

Considering the results presented in Table 6, because of the particle size one can recognize that all ore samples pre-treated with 15% HCl led to almost the same composition. The only noticeable difference corresponds to the presence of CuCl.

If one compares Fig. 7 with Fig. 10, it is possible to observe an identical effect over the surface of the particles, caused by the HCl, which induced cracks in them. The same comparison can be made for Fig. 8 with Fig. 11, where the width and depth of the fractures were clearly amplified by the addition of more concentrated HCl, with a noticeable effect on the copper ion recovery, as evident from Fig. 9. These hints to why particle size is a less relevant factor, as bigger particles were as fractured as smaller ones, solving the problem of possible passivation, and allowing for an efficient chemical dissolution of the relevant metal species from within.

Figure 10 Source: The Authors

Fig. 11 Source: The Authors

In Table 7, ΔG values for some of the reactions discussed in this paper are shown. These values were calculated using the HSC Chemistry software [43].

Table 7

Equation	ΔG (kcal)
5	-185.0
11	-104.1
12	-88.77
13	-23.40
14	-28.83
15	-48.94
16	-96.89
17	-92.12
18	-85.58
19	-72.72
20	-67.96
21	-61.41
22	-11.32
23	-8.145
24	-3.782
25	-56.22
26	-53.07
27	-48.71

Source: The Authors