Surface Physicochemical Property Differences Between Gold-Bearing and Gold-Free Pyrite for Efficient and Clean Processing of Refractory Pyritic Gold Ores

Abstract

Selective separation of gold-bearing pyrite from gold-free pyrite through flotation to improve the gold-to-sulfur ratio in the feed can significantly enhance the throughput of autoclaves, thus achieving efficient and clean processing of refractory pyritic gold ores. To achieve this expectation, this study examined the surface physicochemical differences between gold-bearing and gold-free pyrite under flotation conditions using cyclic voltammetry, polarization curve testing, electrochemical impedance spectroscopy (EIS), X-ray photoelectron spectroscopy (XPS), and density functional theory (DFT) simulations. Electrochemical tests showed higher reactivity in gold-bearing pyrite, with reactivity positively correlated to gold content. XPS results indicated more oxidation products on gold-bearing pyrite surfaces under identical conditions. DFT simulations revealed that the presence of gold reduced the oxygen adsorption energy on the pyrite surface while enhancing interactions between oxygen atoms and sulfur and iron atoms. Based on these findings, the selective separation of gold-bearing and gold-free pyrite in the flotation process can be explored through pulp aeration pre-oxidation combined with collectors demonstrating selectivity toward barren pyrite (e.g., dithiocarbamate collectors). This study provides theoretical foundations for the efficient exploitation and utilization of refractory gold-bearing pyrite resources.

1. Introduction

Gold ores with a gold leaching rate below 80% in the cyanide leaching process are usually referred to as refractory gold ores. In this type of ore, gold is usually locked in sulfides (primarily pyrite and arsenopyrite) as submicron particles or even atomic forms, making it difficult for the leaching agent to access [1,2]. Consequently, the extraction of gold from refractory ores compared to free-milling gold ores requires more advanced technology and higher initial costs [3]. Despite these challenges, refractory gold ores still receive considerable attention due to their vast reserves, higher average gold grades, and the depletion of easily processed free-milling gold resources [4]. Advanced gold processing technologies such as ultrafine grinding, bio-oxidation, roasting, and pressure oxidation have been developed to address these refractory gold ores [5,6,7,8].

Gold is typically extracted from refractory pyritic ores by destroying the pyrite matrix through pressure oxidation, which converts the pyrite into a loose and porous iron oxide, thereby exposing the encapsulated gold for subsequent cyanidation [9]. During pressure oxidation, sulfur in pyrite is oxidized and releases a large amount of heat, which can reduce the pressure and heat artificially applied to the autoclave [10,11]. This phenomenon serves as the basis for the autothermal operation of the autoclave [12]. However, when the sulfur content surpasses a specified threshold, the feed must be diluted to mitigate excessive pressure and heat in the autoclave. Consequently, high sulfur content in the feed significantly reduces the processing capacity of the autoclave, posing a major challenge in the metallurgical production of refractory pyritic gold ores [13,14].

An effective method to address the aforementioned problem is by separating gold-bearing pyrite from gold-free pyrite, thereby increasing the gold-to-sulfur ratio (Au:S ratio) in the autoclave feed [15]. Flotation has always been an efficient and clean beneficiation technology, capable of enriching valuable minerals while discarding gangue [16,17,18]. It exploits the differences in surface physicochemical properties between valuable minerals and gangue to achieve separation [19,20]. However, limited research exists on the differences in surface physicochemical properties between gold-bearing and gold-free pyrite in flotation environments. Chen et al. employed DFT calculations to study the occurrence of gold in pyrite and found that it significantly affected the electronic and structural properties of pyrite [21]. Huai et al. examined the galvanic interaction between gold and pyrite in various slurry environments by coupling gold with pyrite [14,22]. Chang et al. found through DFT calculations that gold doping could enhance the adsorption of xanthate on the pyrite surface [23]. However, extensive industrial data have shown that gold-bearing and gold-free pyrite cannot be selectively separated in a xanthate system. Current research on gold-bearing pyrite is either limited to density functional theory or primarily focuses on pyrite rather than gold-bearing pyrite itself. Therefore, in response to this situation, two pyrite samples with different gold contents were prepared, and the physicochemical property differences between gold-bearing and gold-free pyrite in a flotation context were investigated using cyclic voltammetry, polarization curve testing, electrochemical impedance spectroscopy (EIS), XPS surface analysis, and DFT simulations. This study offers a theoretical basis for the flotation separation of gold-bearing and gold-free pyrite, potentially add ressing the issues of low efficiency and high energy consumption in the metallurgical processing of refractory pyritic gold ores.

2. Material and Methods

2.1. Material and Reagents

The gold-free pyrite sample was purchased from Yunnan Province, China. The gold-bearing pyrite samples were obtained by flotation from a gold-containing pyrite ore in Hunan Province, China. To remove residual flotation reagents from the sample surfaces and further purify the samples, the gold-bearing pyrite was ultrasonically cleaned multiple times with a 1% hydrochloric acid solution, after which the acid-washed samples were transferred to an agate mortar for grinding. Subsequently, the samples were immersed in deionized water for further ultrasonic cleaning. To account for potential surface modifications induced by the above pretreatment methods, identical processing procedures were applied to the gold-free pyrite samples as well. The samples underwent X-ray diffraction (XRD) and chemical analysis to determine their purity and gold content. The XRD results are shown in Figure 1, and the chemical analysis results are presented in

Table 1. Chemical analysis of the three pyrite samples (%).

Sample	Elements
	Au(g/t)	Fe	S	Pb	Zn	O	Si	Al	Others
1	0	45.84	53.40	0.01	0.03	0.32	0.12	0.09	0.19
2	7.58	44.87	50.62	0.08	0.35	1.97	1.33	0.22	0.66
3	29.69	43.97	49.21	0.07	0.33	3.11	2.14	0.34	0.83

. As indicated in Table 1, the purity of the three pyrite samples exceeded 90%. The gold content was 0 g/t in sample 1, 7.58 g/t in sample 2, and 29.69 g/t in sample 3. In subsequent experiments, these three pyrite samples were differentiated by their gold content.

Graphite and paraffin were utilized to fabricate electrode sheets; disodium hydrogen phosphate and sodium dihydrogen phosphate were employed to prepare a buffer solution; potassium nitrate served as a supporting electrolyte; calcium hypochlorite acted as an oxidant; and hydrochloric acid functioned as a pH regulator. All these chemical reagents were of analytical grade. Deionized water was utilized for all experiments.

2.2. Electrochemical Measurements

A Gamry electrochemical workstation was employed for cyclic voltammetry, polarization curves, and electrochemical impedance spectroscopy (EIS) tests. The conventional three-electrode system, comprising an Ag/AgCl reference electrode, a graphite counter electrode with an area of 1 cm², and a mineral carbon paste working electrode, was employed in these electrochemical tests. The preparation process for the mineral carbon paste electrode is as follows: Weigh the appropriate amounts of the mineral, graphite, and paraffin into a beaker, ensuring the mass ratio of 7:2:1 is maintained accurately. After thoroughly mixing the components together to achieve a uniform mixture, proceed to heat the combination over an alcohol lamp. Once the paraffin starts to slightly melt, quickly transfer the mixture into a pre-prepared mold. Press the mixture firmly into the mold to form an electrode with an area approximately measuring 1.0 cm². After the electrode has been formed, proceed to polish the surface of the mineral electrode meticulously using metallographic sandpaper until it achieves a smooth and even finish [19]. Prior to each test, the mineral electrodes were polished with 3000- and 8000-grit metallographic sandpaper to expose a fresh electrode surface. A buffer solution of pH 4.0 phosphate containing 0.1 M potassium nitrate was prepared for each electrochemical test. The polarization curves and impedance spectra were fitted using Gamry Echem Analyst software (7.07). All the potentials reported in this paper have been converted into standard hydrogen electrode (SHE) values.

2.3. X-Ray Photoelectron Spectroscopy

In this study, X-ray photoelectron spectroscopy (XPS) was utilized to characterize the surface oxidation of various pyrite samples. The sample preparation for XPS involved weighing 2 g of pyrite samples (−38 μm) and placing them in a beaker, followed by the addition of 40 mL of a 100 mg/L calcium hypochlorite solution. The slurry’s pH was adjusted to 4.0 and stirred for 20 min. Subsequently, the samples were filtered and placed in a freeze-drying oven. XPS analysis was conducted once the sample was completely dried, utilizing a K-Alpha X-ray photoelectron spectrometer (Thermos Scientific Co., Waltham, MA, USA). During the testing process, Al Kα was used as the sputtering source. The operation was conducted under an accelerating voltage of 16 kV and a current of 14.9 mA, with the pressure in the analysis chamber maintained at 2.31 × 10⁻⁸ mBar. Initially, a survey scan was employed to analyze the surface elements of the three pyrite samples. Subsequently, high-resolution scans were conducted for the C1s, Fe2p, and S2p elements. The C1s peak was utilized for spectral correction to compensate for surface charging effects, whereas the Fe2p and S2p peaks were employed to characterize the surface oxidation of the three pyrite samples. The experimental data were analyzed and fitted using Thermo Advantage software (5.948).

2.4. Calculation Methods

Under the framework of quantum mechanics theory, the adsorption behavior of a single oxygen molecule on different pyrite surfaces was calculated using the CASTEP module in Materials Studio 2018 software [24]. The interactions between core and valence electrons were described by ultrasoft pseudopotentials during the calculations [25]. The exchange-correlation potential was established using the Perdew–Burke–Ernzerhof (PBE) functional under the generalized gradient approximation (GGA). The Kohn–Sham wave functions of the valence electrons were described using a plane-wave basis set with a cutoff energy of 400 eV. A vacuum layer of 20 Å was set above the pyrite 100 surface. The Broyden–Fletcher–Goldfarb–Shanno (BFGS) atomic coordinate optimization algorithm was used to optimize the model. All calculations employed the DFT-D (Grimme) method for van der Waals dispersion corrections. The Hubbard U value for Fe in the d orbitals was set to 2.5 eV [26]. The convergence criteria for geometric optimization were: energy convergence threshold of 1.0 × 10⁻⁵ eV/atom, force convergence threshold of 0.03 eV/Å, stress convergence threshold of 0.02 GPa, displacement convergence threshold of 0.001 Å, and self-consistent iteration convergence accuracy of 1 × 10⁻⁶ eV/atom. A 2 × 2 × 1 Monkhorst–Pack (MP) k-point sampling density was used [27,28]. The valence electrons for each element were as follows: Fe 3d⁶4s², S 3s²3p⁴, O 2s²2p⁴, Au 5d¹⁰6s¹.

To evaluate the impact of the gold occurrence and content on the surface properties of pyrite, three geometric models of pyrite were established, as shown in Figure 2. In Model 1, the pyrite unit cell was expanded into a 2 × 2 × 2 supercell to simulate gold-free pyrite. In Model 2, gold was incorporated into the pyrite structure by substituting sulfur atoms with monovalent gold ions, thus simulating low-gold-content pyrite [21]. In Model 3, gold was embedded into the pyrite structure by forming chemical bonds with surface atoms as gold nanoclusters, simulating high-gold-content pyrite [29]. The calculation of the oxygen adsorption energy is based on the following formula [30]:

$$E_{ads}=E_{{surface/O}_2}-E_{surface}-E_{O_2}$$

where $E_{ads}$ represents the adsorption energy, $E_{{surface/O}_2}$ represents the energy of the different pyrite surfaces with adsorbed oxygen, $E_{surface}$ represents the energy of the different pyrite surfaces, and $E_{O_2}$ represents the energy of the oxygen molecule calculated in a cubic cell.

3. Results and Discussion

3.1. Cyclic Voltammetry Measurement

The oxidation state of the mineral surface is crucial for mineral flotation, and an appropriate surface oxidation state can even enable the collectorless flotation of pyrite [31]. Thus, the influence of gold occurrence on the oxidation properties of the pyrite surface was initially investigated using a cyclic voltammetry test. Figure 3 displays the cyclic voltammetry curves of various pyrite samples in a buffer solution with pH = 4. Two anodic peaks (A₁ and A₂) and two cathodic peaks (C₁ and C₂) were observed in the CV curves of all pyrite samples. The anodic peak A₁, observed at approximately 100 mV, was attributed to the oxidation of pyrite, resulting in the formation of iron hydroxide and elemental sulfur, as illustrated in reaction (1) [14,32]. The anodic peak A₂, commencing at approximately 700 mV, resulted from the oxidation of pyrite into ferric hydroxide and sulfate, as detailed in reaction (2) [33,34].

$${FeS}_2+3H_{2}O\to {Fe\left(OH\right)}_3+2S+3H^{+}+3e^{-}$$

(1)

$${FeS}_2+11H_{2}O\to {Fe\left(OH\right)}_3+2{SO}_4^{2-}+19H^{+}+15e^{-}$$

(2)

The cathodic peak C₁, identified around 320 mV, was attributed to the reduction of iron hydroxide, as outlined in reaction (3) [35]. The broader range and elevated current of cathodic peak C₂ suggested a more complex reduction reaction, typically encompassing the reduction of elemental sulfur (S⁰), polysulfide (FeS_n), metal-deficient sulfide (Fe_1−XS₂), and pyrite [36]. Observations from both the positive and negative sweep processes revealed that auriferous pyrite exhibited higher oxidation and reduction currents than gold-free pyrite, demonstrating more active substances on the surface of auriferous pyrite involved in the reactions, thereby yielding a greater current. Therefore, it can be concluded that auriferous pyrite exhibits higher chemical activity than gold-free pyrite under identical conditions.

$${Fe\left(OH\right)}_3+3H^{+}+3e^{-}\to {Fe}^{2+}3H_{2}O$$

(3)

3.2. Polarization Curve Measurement

It is well-recognized that sulfide ores possess semiconductor properties. When immersed in the solution system, redox reactions occur on the surface of sulfide ores, leading to their corrosion [37]. Consequently, polarization curves are commonly employed to study the self-corrosion behavior of sulfide ores. Figure 4 displays the Tafel curves of various pyrite electrodes in a pH 4.0 buffer solution.

Table 2. Tafel parameters of different pyrite electrodes in pH 4.0 buffer solution.

Sample	Ecorr/V	Icorr/μA	ba/V	bc/V	Rp/Ω
0 g/t	0.44	1.00	0.181	0.138	33,976.06
10 g/t	0.32	1.36	0.176	0.133	24,222.11
30 g/t	0.44	1.93	0.231	0.143	19,823.60

presents the Tafel parameters of various pyrite samples in a pH 4.0 buffer solution, analyzed using Gamry Echem Analyst software. E_corr represents the corrosion potential. I_corr denotes the corrosion current, which is commonly used to characterize the rate of the corrosion reaction. A higher corrosion current indicates a faster reaction rate. While b_a and b_c represent the Tafel constants of the anode and cathode, respectively. R_p denotes polarization resistance, inversely proportional to the corrosion rate and calculated using the Stern–Geary equation [38]:

$$R_p={\displaystyle \frac{b_a{b}_c}{2.303(b_a+b_c)I_{corr}}}$$

According to Table 2, the I_corr of gold-free pyrite was the lowest at 1.00 μA, followed by pyrite with a gold content of 10 g/t at 1.36 μA, and the highest I_corr of 1.93 μA was observed in pyrite with a gold content of 30 g/t. The higher corrosion current of auriferous pyrite indicated a faster corrosion reaction rate. The polarization resistance of the gold-free pyrite was 33,976.06 Ω, while the polarization resistances for the pyrites with gold contents of 10 g/t and 30 g/t were 24,222.11 Ω and 19,823.60 Ω, respectively. Compared to gold-free pyrite, auriferous pyrite exhibited lower polarization resistance, suggesting a smaller reaction resistance. Thus, the faster corrosion rate and smaller reaction resistance suggest that auriferous pyrites possess heightened reactivity, which escalates with increasing gold content. This is consistent with results from cyclic voltammetry tests.

3.3. Electrochemical Impedance Spectroscopy

Due to the semiconducting nature of pyrite, it can form a double-layer structure at the interface with the aqueous solution in flotation environments. Therefore, electrochemical impedance spectroscopy (EIS) was used to characterize the parameters related to the double-layer structure of pyrite to investigate the effect of gold presence on the surface properties of pyrite.

Electrochemical impedance spectroscopy (EIS) diagrams for various pyrite electrodes in a pH 4.0 buffer solution are depicted in Figure 5. As shown in Figure 5a, the impedance modulus of the three gold-bearing pyrite electrodes remained constant in the high-frequency region, indicating the possible presence of solution resistance in the system. Figure 5b presents the phase angle versus frequency relationship, where negative phase angles were observed across most frequency ranges—a characteristic feature of capacitive behavior. Notably, in the low-frequency region, the phase angles of all three pyrite electrodes approached −π/4, suggesting the existence of Warburg impedance. The transition to positive phase angles in the high-frequency region implies current lagging behind voltage, typically caused by inductive behavior, possibly due to inductive impedance in the system. The Nyquist plots in Figure 5c further revealed a linear correlation between the imaginary and real parts of impedance in the low-frequency range, providing additional evidence for the presence of Warburg impedance in the system.

Based on the aforementioned information, the equivalent circuit for the system was determined to be LR_s(Q(R_ctZ_w)), as shown in Figure 5d. In the equivalent circuit diagram, LR_s(Q(R_ctZ_w)), L represents system inductance, Rs denotes solution resistance, Rct stands for charge transfer resistance, Q signifies the constant phase element, and Z_w corresponds to Warburg impedance. R_ct represents the difficulty of charge transfer across the electrode surface during the transition at the solid–liquid interface. Q acts as a non-ideal substitute for the double-layer capacitor C at the electrode/electrolyte interface. The parameter n for Q ranges from 0 to 1, where Q becomes the ideal capacitor C at n = 1 [34,39]. The measured data closely matched the fitted data shown in Figure 5, and the low chi-square value (

Table 3. The electrical parameters obtained from the equivalent circuit fitting.

pH	Mineral	L (10−7 H)	Rs (Ω/cm2)	Rct (kΩ/cm2)	Q (10−5 Ssn/cm2)	n	W (103 Ω−1S0.5cm2)	χ (10−3)
4	0 g/t	6.14	8.00	1.01	21.67	0.69	7.13	1.86
	10 g/t	5.83	10.51	0.52	21.15	0.66	1.14	1.69
	30 g/t	5.81	11.58	0.15	9.26	0.77	6.78	1.72

) suggested the appropriateness of the chosen equivalent circuit.

Table 3 displays the electrical parameters derived from the equivalent circuit modeling. The similar R_s values among the three pyrite samples indicated stability in the electrical properties of the buffer solution throughout the measurement period. Table 3 revealed a distinct inverse correlation between gold content and charge transfer resistance in pyrite electrodes. Specifically, the charge transfer resistance measured 1.01 kΩ/cm² for gold-free pyrite, decreasing to 0.52 kΩ/cm² for 10 g/t gold-bearing pyrite, and further dropping to 0.15 kΩ/cm² for 30 g/t gold-bearing pyrite. This systematic reduction in charge transfer resistance with increasing gold content (from 0 to 30 g/t) demonstrates that gold occurrence significantly enhances the charge transfer kinetics at the pyrite–electrolyte interface. Given that electrode reactivity is inversely proportional to charge transfer resistance, lower R_ct values indicate increased reactivity in auriferous pyrite, aligning with results from cyclic voltammetry and polarization curve tests.

3.4. X-Ray Photoelectron Spectroscopy Analysis

The electrochemical experimental results indicated that gold-bearing pyrite exhibited higher chemical activity than gold-free pyrite, particularly evident in the cyclic voltammetry findings where more oxidation products formed on the surface of auriferous pyrite at the same potential. To validate these findings, both gold-bearing and gold-free pyrite samples were subjected to identical oxidation conditions, followed by the use of XPS to characterize the oxidation on various pyrite surfaces.

Figure 6a–c display the XPS high-resolution spectra of S2p on various oxidized pyrite surfaces. The peak positions of different sulfur species across the surfaces of the three samples showed consistency. The double peaks in the range of approximately 160.5–162.5 eV correspond to the split peaks of surface monosulfide (S²⁻) [40,41]. The double peaks in the range of approximately 161.5–163.5 eV are attributed to the split peaks of disulfide (S₂²⁻) within the bulk pyrite [42]. The double peaks in the range of approximately 163.5–165.5 eV are indicative of the split peaks of polysulfides or elemental sulfur (S_n/S_n²⁻), intermediate products of sulfur oxidation on the pyrite surface [43]. The double peaks within the range of approximately 168.0–170.0 eV correspond to the split peaks of sulfate (SO₄²⁻), the final product of sulfur oxidation [44].

Table 4. The distribution of S2p species on different oxidized pyrite surfaces.

Species	0 g/t		10 g/t		30 g/t
	B.E.	at. %	B.E.	at. %	B.E.	at. %
S2−2p1/2	161.56	7.49	161.63	6.75	162.16	5.92
S2−2p3/2	160.92	3.86	161.13	10.46	161.45	11.64
S22−2p1/2	163.02	23.38	163.03	25.39	163.30	26.24
S22−2p3/2	161.89	47.03	161.98	38.43	162.18	34.57
Sn/Sn2−2p1/2	165.26	4.39	165.19	4.58	166.07	3.17
Sn/Sn2−2p3/2	163.75	9.03	163.81	7.55	164.48	6.83
SO42−2p1/2	169.31	0.82	169.46	1.56	169.60	2.39
SO42−2p3/2	168.01	4.02	168.23	5.28	168.35	9.24

exhibits the binding energy (B.E.) and relative atomic content (at. %) of the peaks for the aforementioned sulfur species. SO₄²⁻ is the final product of sulfur oxidation on the pyrite surface, and its concentration percentage content offers valuable insights into the oxidation state of the pyrite surface. The SO₄²⁻ concentration on the gold-free pyrite surface was 4.84%, while the SO₄²⁻ concentrations for pyrite samples containing 10 g/t and 30 g/t gold were 6.84% and 12.14%, respectively. Based on the above data, it is evident that the degree of oxidation of the pyrite surface increases with the gold content.

Figure 6d–f display the XPS high-resolution spectra of Fe2p_3/2 on various oxidized pyrite surfaces. The XPS spectra of Fe2p_3/2 were characterized by peaks at Fe (706.4 eV) from bulk pyrite, Fe (707.7 eV) from the pyrite surface, and hydroxylated iron Fe-O/OH (710.8 eV) [45,46,47]. Hydroxylated iron Fe (III)-O/OH represents an oxidation product of Fe (II) on the pyrite surface. Following oxidation under identical conditions, the hydroxylated iron content was 9.86% on gold-free pyrite surfaces, 11.91% on surfaces of pyrite with 10 g/t gold, and highest at 19.14% on surfaces with 30 g/t gold, as shown in

Table 5. Distribution of Fe2p3/2 species on different oxidized pyrite surfaces.

Species	0 g/t		10 g/t		30 g/t
	B.E.	at. %	B.E.	at. %	B.E.	at. %
Bulk Fe(II)-S	706.32	56.76	706.43	55.84	706.59	55.41
Fe(II)-S	707.72	33.39	707.68	32.34	707.74	25.46
Fe(III)-O/OH	710.80	9.86	711.01	11.81	710.80	19.14

.

Under identical oxidation conditions, gold-bearing pyrite yielded more oxidation products and, consequently, higher redox currents, aligning perfectly with the results of cyclic voltammetry. Thus, the XPS results validated the electrochemical experiments’ conclusion that gold-bearing pyrite exhibits higher chemical activity than gold-free pyrite, with its activity escalating as gold content increases.

3.5. DOS Analysis of Different Pyrite Surfaces Before and After Oxygen Adsorption

Electrochemical tests and XPS analysis results indicate that gold presence significantly alters the electrochemical reactivity of pyrite. To further investigate the impact of gold presence on the surface properties of pyrite, DFT simulations were used to construct adsorption configurations of oxygen on pyrite surfaces with varying gold content, as shown in Figure 7.

The partial density of states (DOS) of iron, oxygen, and sulfur atoms on different pyrite surfaces before and after oxygen adsorption is shown in Figure 8. The Fermi level (Ef) is set to 0 eV. After oxygen adsorption, the DOS of all three atoms exhibits significant changes, indicating a substantial impact of oxygen adsorption on the surface states of the mineral. In all three models, the 2p orbital DOS of oxygen atoms displays a continuous distribution in the energy range of −8.5 to 4 eV post-adsorption, suggesting enhanced electron delocalization. In Model 1, the 3p DOS peak of sulfur atoms below the Fermi level shifts towards lower energy, with a decrease in density of states within the energy range of −5 to −2 eV and an increase in density of states within range of −8 to −6 eV; the 3d orbital DOS of iron atoms shifts towards lower energy, with a decrease in density of states within the energy range of −2 to 0 eV. In Model 2, the 3p orbital DOS of sulfur atoms shifts approximately 1 eV towards lower energy after oxygen adsorption, with a significant decrease in density of states within the range of −6 to −2 eV and a notable increase around −8 eV; the 3d orbital DOS of iron atoms shifts 1 eV towards lower energy, with a significant decrease in density of states within the range of −2 to 0 eV. In Model 3, the 3p orbital DOS of sulfur atoms shifts approximately 0.6 eV towards lower energy after oxygen adsorption, with the DOS peak around −6 eV disappearing; the 3d orbital DOS peak of iron atoms weakens within −2 to 0 eV. Notably, in all models, the 3p orbitals of sulfur atoms and the 3d orbitals of iron atoms overlap extensively with the 2p orbitals of oxygen atoms in the bonding region below the Fermi level, indicating the formation of stable S-O and Fe-O bonds. This is consistent with the detection of sulfur oxides and iron oxides on the oxidized surfaces of the three pyrites in the XPS analysis.

3.6. Adsorption Energy and Structure of Oxygen on Different Pyrite Surfaces

Density of states analysis indicated that oxygen formed stable chemical bonds with sulfur and iron atoms on the pyrite surface after adsorption. Therefore, the relevant parameters of S-O bonds and Fe-O bonds on different pyrite surfaces in the three models were analyzed to investigate the influence of gold occurrence on the surface properties of pyrite. In Model 1, the O-S bond length was 0.1684 nm and in Models 2 and 3, the O-S bond lengths were 0.1667 nm and 0.1618 nm, respectively. The covalent bond radius sum of oxygen and sulfur atoms is 0.175 nm [48], which is greater than the O-S bond lengths in all models, indicating that oxygen atoms formed bonds with the sulfur atoms on the pyrite surface in all models. In Models 1, 2, and 3, the O-Fe bond lengths were 0.1977 nm, 0.1872 nm, and 0.1798 nm, respectively. The covalent bond radius sum of oxygen and iron atoms is 0.2010 nm [49], which is greater than the O-Fe bond lengths in all models, indicating that oxygen atoms also formed bonds with the Fe atoms on the pyrite surface in all models. Notably, as the gold content increased in the models, the O-S and O-Fe bond lengths gradually decreased. The adsorption energies of oxygen on the gold-free pyrite surface in Model 1 were −63.50 kJ/mol, on the low-gold pyrite surface in Model 2 were −68.77 kJ/mol, and on the high-gold pyrite surface in Model 3 are −84.18 kJ/mol. These observations indicated that the presence of gold enhances oxygen adsorption, and as the gold content increases, the adsorption energy of oxygen on the pyrite surface further decreases.

Table 6. Mulliken bond population of S-O bond and Fe-O bond on different pyrite surfaces.

Adsorption Model	Bond	Population
1	S-O	0.22
	Fe-O	0.29
2	S-O	0.23
	Fe-O	0.38
3	S-O	0.29
	Fe-O	0.35

presented the Mulliken bond population of S-O and Fe-O bonds on pyrite surfaces with different gold contents after oxygen adsorption. Compared to the gold-free pyrite in Model 1, the Mulliken bond populations of S-O and Fe-O bonds on the low-gold pyrite surface in Model 2 increased to 0.23 and 0.38, respectively, while on the high-gold pyrite surface in Model 3, they increased to 0.29 and 0.35, respectively. This indicated that the presence of gold enhanced the covalent nature of S-O and Fe-O bonds, thereby strengthening the interactions between oxygen atoms and both sulfur and iron atoms. Therefore, based on the above information, it is evident that the presence of gold reduces the adsorption energy of oxygen on the pyrite surface, while simultaneously shortening the S-O and Fe-O bond lengths and enhancing their covalent character.

4. Conclusions

The physicochemical differences between gold-free pyrite, low-gold pyrite, and high-gold pyrite using cyclic voltammetry, polarization curve testing, electrochemical impedance spectroscopy (EIS), XPS surface analysis, and DFT simulations in this study. The main conclusions are as follows:

(1)

All electrochemical analyses indicated that the occurrence of gold enhanced the electrochemical reactivity of pyrite, with reactivity positively correlated with gold content.

(2)

XPS results confirmed the findings of the electrochemical tests, demonstrating that under identical oxidation conditions, high-gold pyrite had the most oxidation products on its surface, followed by low-gold pyrite, and gold-free pyrite had the least.

(3)

DFT simulations revealed that the presence of gold lowered the adsorption energy of oxygen on the pyrite surface while also promoting the interaction between oxygen atoms and sulfur and iron atoms on the pyrite surface.

Based on the differences in electrochemical activity between gold-bearing and gold-free pyrite, their selective separation in the flotation process can be achieved by pulp aeration-induced preferential generation of hydrophobic products on gold-bearing pyrite surfaces combined with selective collector for barren pyrite (e.g., dithiocarbamate collectors). Therefore, this study provides a theoretical foundation for the flotation separation of gold-bearing and gold-free pyrite, which may help resolve the challenges of low efficiency and high energy consumption in the metallurgical processing of refractory pyritic gold ores.