Enrichment of Copper and Cobalt from Pyrite Cinder via Cyclic Leaching: Schwertmannite-Type Iron Precipitation at Low pH

Abstract

Effective iron removal from highly acidic leachates without neutralizer addition is a long-standing challenge in hydrometallurgy. For Cu-Co-bearing pyrite cinder, we demonstrate that a cyclic leaching process achieves this outcome. Operating entirely without neutralizers, the process consumed 96% of residual acid, enriched Cu and Co 4.76- and 3.88-fold, and removed 48.2% of iron, all while maintaining pH below 1.5 across four closed-loop cycles. Thermodynamic analysis reveals that Fe³⁺-${\mathrm{SO}}_4^{2-}$ complexation suppresses free Fe³⁺ to approximately 10⁻⁸ mol/L, ruling out conventional hydrolytic precipitation and directing precipitation towards schwertmannite-type basic ferric sulfate (Fe₈O₈(OH)₆SO₄). The formation of this sulfate-bearing Fe precipitate is consistent with the thermodynamic analysis, the iron mass balance, and a 176% increase in solid-phase sulfur. This neutralizer-free strategy offers a sustainable paradigm for recovering critical metals from iron-rich secondary resources.

1. Introduction

Copper- and cobalt-bearing pyrite is a primary feedstock for the production of strategic metals Cu and Co. Conventional extraction routes typically involve oxidation and leaching [1], including oxidative roasting–acid leaching [2,3], sulfation roasting–leaching [4,5], and bioleaching [6,7]. Pyrite cinder (PyC), an industrial byproduct of sulfuric acid manufacturing, is classified as a high-potential secondary resource containing abundant Fe together with associated Cu, Co, and Ni [8]. Approximately 12 million tonnes of PyC are produced annually in China [2], with iron oxides as the dominant phase and recoverable contents of Cu, Co, and Ni, offering strong economic and environmental incentives for valorization.

The challenge of selectively recovering critical metals (Cu, Co, Ni) from Fe-rich secondary matrices, including pyrite cinder, red mud, steelmaking slag, and metallurgical residues, represents a common bottleneck across the hydrometallurgy industry. In the specific context of PyC, extensive research has thus focused on the selective separation of Cu and Co from the iron matrix in PyC [9]. During roasting, Co is primarily oxidized to higher-valence states such as Co(III) species with low acid solubility, requiring reducing conditions for effective dissolution [10,11]. Conventional reductive leaching strategies therefore employ exogenous reductants such as sodium metabisulfite or SO₂ to improve Co dissolution [12,13]. Despite achieving higher recoveries, these conventional reductive leaching strategies carry intrinsic drawbacks. They increase operational costs, pose environmental hazards from SO₂ emissions, and can reduce Cu-Co separation selectivity through Cu co-precipitation under excess reductant dosage.

To address the limitations of external reductants, endogenous reductive acid leaching has emerged as a promising green strategy that utilizes the intrinsic reducing potential of the minerals. Yao et al. [14] demonstrated that co-blending a reductive ore with oxidized Cu-Co ore significantly enhanced Co extraction from 22% to 76% without additional reductant. This result confirms that controlling redox potential through endogenous reduction effectively intensifies the leaching of high-valence Co.

In parallel, reductive roasting pre-treatment followed by leaching has been adopted to further improve Cu and Co recovery. Zhang et al. [15] applied a reductive roasting–leaching–magnetic separation process to treat PyC. This process achieved 82.18% Cu leaching efficiency and produced high-grade iron concentrate, realizing the synergistic recovery of Cu and Fe. Yu et al. [16] employed reductive roasting to reconstruct the mineral phases in PyC, converting refractory Co-bearing phases into readily leachable Co(II). This approach yielded maximum leaching efficiencies of 86.15% for Cu and 79.61% for Co.

Despite the relatively high leaching efficiencies achieved in roasting–acid leaching systems, several critical bottlenecks persist. First, the leachate exhibits high acidity, with residual acid concentrations reaching up to 60 g/L, rendering it poorly compatible with downstream solvent extraction processes for Cu. Second, downstream solvent extraction typically requires a pH near 1.5, yet conventional neutralization to achieve this introduces calcium and magnesium impurities [17]. Third, Fe³⁺ hydrolysis and co-precipitation readily occur, causing losses of Cu and Co and hindering selective removal of Fe at low pH. In strongly acidic sulfate media (pH < 2), iron is conventionally precipitated not as hydrolytic Fe(OH)₃ but as basic ferric sulfates, namely schwertmannite and jarosite; yet achieving selective Fe removal by this route without co-precipitating Cu and Co remains a persistent challenge.

To reduce acid consumption, improve leaching efficiency, and enhance process sustainability, cyclic leaching has been widely adopted as a process optimization strategy [18]. Ntakamutshi et al. [19] demonstrated that cyclic leaching significantly reduces unreacted SO₂ emissions, lowers acid consumption, and increases recovery of Cu and Co. However, existing cyclic leaching studies for PyC focus almost exclusively on reductant utilization efficiency and overall leaching performance, with no systematic investigation of Fe valence evolution, phase transformation during cycling, or their implications for selective Cu-Co separation.

Consequently, simultaneously achieving residual acid consumption, Cu-Co enrichment, and the selective removal of Fe in an acidic system—without neutralizing agents or external reductants—represents a persistent and critical challenge in PyC valorization. Preliminary experiments, however, revealed an anomalous phenomenon: efficient iron removal occurred under strongly acidic conditions (pH < 1.5) without compromising the recovery of target metals, a finding that cannot be explained by conventional hydrolytic precipitation.

To investigate the mechanism behind this anomaly and its potential to address the aforementioned challenge, this study proposes a neutralizer-free cyclic leaching strategy. The process was systematically evaluated for its feasibility to synergistically consume residual acid, enrich Cu and Co, and remove iron. Based on these findings, an optimized cyclic leaching process was developed. This work thus establishes a theoretical framework for the efficient recovery of valuable metals from iron-rich secondary resources.

2. Materials and Methods

2.1. Materials

The Cu-Co-bearing PyC used in this study was obtained from a metallurgical plant in Inner Mongolia, China. Analytical-grade sulfuric acid (H₂SO₄) was obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Deionized water was used throughout the experiments.

2.2. Experimental Procedures

The cyclic leaching procedure is illustrated in Figure 1.

(a)

Primary Leaching: Initial leaching was conducted using a 100 g/L H₂SO₄ solution at a liquid-to-solid ratio of 3:1, 90 °C for 2 h. Upon completion of the reaction, the slurry was hot-filtered to yield filtrate 1 and residue 1. The pH and the concentrations of Cu, Co, total iron (TFe), and Fe²⁺ in Filtrate 1 were measured.

(b)

Cyclic Leaching: The filtrate from the preceding leaching step was reused as the leaching agent in subsequent cycles without the addition of fresh acid or neutralizing agents. In each cycle, fresh PyC was subjected to leaching at a liquid-to-solid ratio of 1:1 at 90 °C for 2 h. After each cycle, the resulting filtrate and residue of that cycle (denoted filtrate n and residue n, n = 1–4) were separated, and the pH of the filtrate was recorded. Cu, Co, TFe, and Fe²⁺ concentrations were also analyzed. Cyclic leaching was continued until the pH of the leaching solution exceeded 1.0, which was defined as the endpoint of the process.

The closed-loop system was specifically designed to enable stepwise reuse of the leachate. This design facilitates continuous enrichment of Cu²⁺ and Co²⁺, in situ consumption of residual acid, and directional migration and removal of Fe. All residues were hot-filtered and thoroughly washed with deionized water before drying to remove retained leachate and soluble sulfate salts.

2.3. Characterization

Elemental concentrations of Cu, Co, and other metals in the leaching solutions were determined using an inductively coupled plasma optical emission spectrometer (ICP-OES; Agilent 5110, Mulgrave, Australia). TFe and Fe²⁺ contents in both the PyC and cyclic leaching residues were quantified by acid digestion followed by titration with potassium dichromate. Sulfur in the raw material was analyzed using a carbon-sulfur analyzer (Leco CS744, St. Joseph, MI, USA). For the determination of Fe²⁺, an aliquot of the hot-filtered, H₂SO₄-acidified filtrate was taken immediately after sampling, kept under a nitrogen blanket, and titrated within 15 min with standard K₂Cr₂O₇ solution using sodium diphenylamine sulfonate as indicator in an H₂SO₄–H₃PO₄ medium; TFe was determined on a separate aliquot after reduction of Fe³⁺ (SnCl₂–TiCl₃ reduction), and Fe³⁺ was obtained by difference. All reported values are the means of triplicate determinations.

The residual H₂SO₄ concentration of each filtrate was determined by acid–base titration of an aliquot with standardized NaOH solution to a pH-meter endpoint of 4.0, with a correction for the acid consumed by the hydrolysis of Fe³⁺ and Al³⁺; each reported value is the mean of three titrations. Phase composition of the PyC and cyclic leaching residues was examined by X-ray diffraction (XRD; Rigaku Ultima IV, Takatsuki, Japan). Surface chemical states in the cyclic leaching residues were investigated using X-ray photoelectron spectroscopy (XPS; Thermo Scientific K-Alpha, Waltham, MA, USA), which revealed changes in Fe speciation during processing. The XPS spectra were fitted using XPSPeak 4.1. Morphological and elemental distributions of PyC were characterized by scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDS; Zeiss EVO18 and Quantax 400, Berlin, Germany). The relative atomic ratios and spatial distribution of Cu, Co, Fe, S, and O were obtained.

3. Results

3.1. Material Composition

Table 1. Main elemental composition of Cu-Co-bearing pyrite cinder (PyC).

Elements	S	TFe	Cu	Fe2+	Co	Si	Ca	Al	Mg
Content/wt%	0.13	63.87	0.82	1.37	0.36	1.11	0.96	0.54	0.67

shows the main elemental composition of the Cu-Co-bearing PyC. Iron is the predominant element, with a TFe content of 63.87 wt%. Fe²⁺ constitutes 1.37 wt%, indicating the presence of appreciable low-valence Fe species. Copper and Co are present at 0.82 wt% and 0.36 wt%, respectively, reflecting significant combined economic potential for recovery. Total sulfur content is 0.13 wt%, indicating incomplete sulfur removal during roasting, with residual sulfide or sulfate phases that may act as endogenous reducing agents and complexation sites in the leaching system. Minor impurities including Ca, Mg, Al, and Si are also detected, consistent with typical PyC composition.

3.2. SEM-EDS Analysis

SEM-EDS was employed to examine the morphology, mineralogical distribution, and elemental composition of the PyC. As shown in Figure 2, PyC is oxidized and dominated by iron oxide phases. Composite regions composed of Si, Ca, Mg, Al, Fe, and other elements are also present. SEM-EDS spot analyses (

Table 2. Representative SEM-EDS spot analyses (wt%) of the Cu-Co-bearing pyrite cinder.

Spot	O	Mg	Al	Si	S	Ca	Fe	Co	Cu	Phase Assignment
1	23.85	0.85	2.37	2.29	0.17	0.24	67.08	1.01	1.71	Fe–Si–O (fayalite-type)
2	25.82	1.83	1.16	0.98	0.15	0.15	69.04	0.32	0.56	Fe–Mg–Al–O (magnesioferrite-type)
3	44.96	0.60	2.10	19.48	0.38	27.18	3.97	0.70	0.64	Ca–Si–O (wollastonite-type)

) confirm the heterogeneity of the cinder: the dominant spots are Fe–O-rich (iron oxide matrix, Fe > 67 wt%), together with Fe–Si–O regions (fayalite-type), Fe–Mg–Al–O regions (magnesioferrite-type) and Ca–Si–O regions (wollastonite-type). Based on the combined SEM-EDS evidence and literature analysis [11], the additional phases present in the system include fayalite (Fe₂SiO₄), magnesioferrite (MgFe₂O₄), wollastonite (CaSiO₃), and calcium sulfate (CaSO₄); each assignment is supported by the corresponding spot composition, and CaSO₄ is consistent with the residual sulfur content of the cinder.

3.3. Cyclic Leaching—pH and Acid Evolution

Figure 3 illustrates the evolution of filtrate pH and residual H₂SO₄ concentration during cyclic leaching. The pH exhibited a monotonic stepwise increase from an initial value of 0.20 after primary leaching to 1.31 after 4 cycles, with a corresponding 96% reduction in residual H₂SO₄ concentration from 60.00 g/L to 2.40 g/L. This in situ acidity reduction adjusts the leachate pH to the range of 1.2–1.5 required for downstream Cu solvent extraction, eliminating the need for neutralization pre-treatment and the associated introduction of impurities. This net H⁺ consumption is attributed to the continuous acid dissolution of basic metal oxides and readily leachable Cu/Co-bearing phases in each newly added batch of PyC, which overcomes the H+ generated from secondary reactions, such as the oxidation of residual sulfides by Fe³⁺.

3.4. Copper and Cobalt Enrichment

Dissolved Cu and Co concentrations exhibited cumulative enrichment throughout the cyclic leaching process, as presented in Figure 4. Because a fresh batch of PyC is dissolved into the closed-loop solution in every cycle while a fraction of the solution volume is retained in the filter cake, the concentration increase is a cumulative enrichment effect of the closed loop rather than an enhancement of the per-batch extraction. The Cu concentration increased from 1.98 g/L after primary leaching to 9.42 g/L after 4 cycles, while the Co concentration increased from 0.40 g/L to 1.55 g/L, corresponding to enrichment factors of 4.76 and 3.88, respectively. This enrichment is achieved without solvent extraction, concentration or evaporative treatment and eliminates the neutralization and hydrolytic Fe removal steps employed in conventional processes. Accordingly, co-precipitation losses of Cu and Co associated with Fe hydroxide precipitation are avoided.

The amounts of Cu, Co, and Fe dissolved from each successive batch of PyC were determined from the measured filtrate concentrations and volumes and the known PyC composition (Table 1), and are reported in

Table 3. Per-cycle dissolution of Cu, Co, and Fe from each successive batch of pyrite cinder.

Cycle	Cu Dissolved (g)	Cu Dissolved (%)	Co Dissolved (g)	Co Dissolved (%)	Fe Net Change (g)
1	0.99	38	0.15	13	+1.31
2	0.61	31	0.09	11	−0.74
3	0.29	20	0.05	7	−0.20
4	0.02	2	<0.01	<1	−0.32

. Cu and Co dissolution occurred in every cycle, with the dissolved amount tracking the residual acidity (Figure 3): 0.99, 0.61, 0.29, and 0.02 g Cu (38%, 31%, 20%, and 2% of the Cu introduced with each batch), and 0.15, 0.09, 0.05, and 0.004 g Co (13%, 11%, 7%, and <1%), for Cycles 1–4. The Cu and Co dissolved in Cycles 1–3 account for 96% and 98% of the total accumulated in solution; the fourth cycle contributed primarily to residual acid consumption and iron removal. For Fe, dissolution and precipitation occur simultaneously; the net change was +1.31 g (Cycle 1, net dissolution), −0.74 g (Cycle 2), −0.20 g (Cycle 3), and −0.32 g (Cycle 4). A single Fe leaching efficiency is not defined, and the Fe removal is quantified separately in Section 3.5.

3.5. Iron Dissolution and Removal

Notably, the TFe concentration in the leachate exhibited an initial increase followed by a gradual decrease (Figure 5), reflecting two distinct stages: rapid Fe dissolution, followed by directional Fe removal. This observed decrease in TFe concentration occurred despite the persistently low pH (<1.5) throughout the cyclic process. This observation contrasts with the well-established mechanism of hydrolytic Fe precipitation, which typically requires a pH > 2.5, thus suggesting an alternative iron removal pathway is operative under these conditions.

During the first cycle, the strongly acidic environment promoted rapid dissolution of soluble Fe oxides and Fe phases encapsulated in porous CaSO₄ matrices, increasing the TFe concentration from 5.60 g/L after primary leaching to 9.69 g/L.

From cycles 2 to 4, as the solution pH increased from 0.45 to 1.31 and the residual H₂SO₄ concentration declined, reductive species (primarily unoxidized sulfides and low-valence metal oxides) in fresh PyC gradually dissolved, increasing the reducing capacity of the system. This redox shift facilitated directional Fe removal, with the TFe concentration decreasing steadily to 2.90 g/L after 4 cycles, corresponding to an Fe removal efficiency of 48.2% relative to the primary leachate: ηFe = (TFe after primary leaching − TFe after Cycle 4)/TFe after primary leaching × 100% = (5.60 − 2.90)/5.60 × 100%; relative to the maximum concentration at Cycle 1, the removal is (9.69 − 2.90)/9.69 × 100% = 70.1%. Because fresh PyC is introduced in every cycle, these values are operational metrics referenced to the defined solution states, not mass-based removal efficiencies relative to the cumulative Fe input. The complete Fe balance of the cyclic campaign—covering the solution, the residues, the liquid retained in the filter cakes, and the wash water—is given in

Table 4. Process mass balance of the cyclic leaching campaign (solution, residue, retained liquid, and wash water).

Cycle	pH	TFe (g/L)	PyC Mass (g)	Leachant Vol. (mL)	Residue Mass (g)	Retained Liquid (mL)	Residue Fe (wt%)	Residue Cu (wt%)	Residue Co (wt%)
Primary	0.20	5.60	115.0	345.0	109.8	25.0	65.1	0.24	0.25
1	0.45	9.69	320.0	320.0	311.5	75.0	65.2	0.52	0.32
2	0.80	7.50	245.0	245.0	241.0	70.0	65.2	0.58	0.33
3	1.05	5.00	175.0	175.0	171.9	55.0	65.2	0.66	0.34
4	1.31	2.90	120.0	120.0	118.0	40.0	65.2	0.82	0.36

; the balance closes to within ±5% for every cycle, confirming that the decrease in dissolved Fe corresponds to Fe retained in the solid phase rather than to sampling or washing losses. These results confirm that efficient, stable Fe removal can be achieved under strongly acidic conditions without neutralizer addition.

3.6. Fe²⁺ Evolution

Analysis of the Fe²⁺ concentration and its fraction in TFe shows that the Fe²⁺/TFe ratio increased monotonically with successive leaching cycles. This increase is driven by two concurrent processes: (i) continuous supply of Fe²⁺ through acid dissolution of Fe²⁺-bearing phases—fayalite-type Fe₂SiO₄, magnesioferrite-type spinels (Table 2), and residual low-valence Fe species—in each fresh batch of PyC (1.37 wt% Fe²⁺, Table 1; at L/S = 1:1, up to approximately 1.4 g/L Fe²⁺ per cycle); and (ii) selective removal of Fe³⁺ by schwertmannite-type precipitation (Section 4.4). The dissolved Fe³⁺ concentration decreased from 8.2 g/L (Cycle 1) to 0.49 g/L (Cycle 4), while Fe²⁺ remained in solution, reaching 2.40 g/L (82.8% of TFe) at Cycle 4. The joint effect of sustained Fe²⁺ input and progressive Fe³⁺ removal effectively alleviates the inhibitory effect of Fe³⁺ on Cu and Co dissolution, supporting sustained enrichment of target metals. The mechanisms underlying the simultaneous Fe removal under these conditions are discussed in detail in Section 4.

3.7. Sulfur Evolution in the Solid Phase

S 2p XPS analysis was conducted to track sulfur evolution across leaching cycles, with relative sulfur content quantified via integrated peak area (

Table 5. Relative S 2p peak area of PyC and cyclic leaching residues.

Sample	Relative Peak Area of Sulfur
PyC	19,695
Leached Residue 0	5375
Residue 1	11,072
Residue 2	11,302
Residue 3	12,664
Residue 4	14,851

). The results indicate that the PyC initially contained the highest sulfur content (relative peak area: 19,695). Following the first cycle, the sulfur content in the residue decreased sharply (to 5375), which is directly attributable to the high initial acid concentration facilitating the dissolution of soluble sulfate species. Notably, with increasing cycle number, the sulfur content in the solid residue increased monotonically, from Residue 1 (11,072) to Residue 4 (14,851), corresponding to a 176% increase relative to the Residue 0 minimum. This progressive sulfur accumulation in the solid phase suggests the formation of a sulfur-bearing precipitate. Detailed mechanistic analysis of this observation is presented in Section 4. It should be noted that Residues 1–4 are separate solids obtained from successive batches of fresh PyC leached under progressively milder solution conditions (pH 0.45 → 1.31); they do not represent transformation stages of a single solid. The monotonic sulfur increase across these successive batches therefore reflects the increasing tendency of the solution phase to deposit sulfur-bearing material as the pH rises, rather than solid-state accumulation within one sample.

4. Discussion

4.1. Thermodynamic Analysis of Fe Speciation

In the highly acidic sulfate system ([${\mathrm{SO}}_4^{2-}$] = 1 mol/L, [H⁺] = 0.05–0.63 mol/L), Fe³⁺ forms strong complexes with sulfate anions [20]:

$${\mathrm{Fe}}^{3+}+{\mathrm{SO}}_4^{2-}\to {\mathrm{Fe}\mathrm{SO}}_4^{+} \log {\mathit{\beta }}_1=4.04 ({\mathit{\beta }}_1\approx \mathrm{11,000})$$

(1)

$${\mathrm{Fe}}^{3+}+2{\mathrm{SO}}_4^{2-}\to {\mathrm{Fe}\left(\mathrm{S}{\mathrm{O}}_4\right)}_2^{-} \log {\mathit{\beta }}_2=5.38 ({\mathit{\beta }}_2\approx \mathrm{240,000})$$

(2)

The side-reaction coefficient α(SO₄) for Fe³⁺ under these conditions is:

$$\mathit{\alpha }\left({\mathrm{SO}}_4\right)=1+{\mathit{\beta }}_1\left[{\mathrm{SO}}_4^{2-}\right]+{\mathit{\beta }}_2[{\mathrm{SO}}_4^{2-}{]}^2\approx \mathrm{251,000}$$

(3)

Consequently, free Fe³⁺ represents only ~0.0004% of total dissolved Fe(III) and all supersaturation calculations must be based on the free Fe³⁺ concentration, corrected for activity effects.

Saturation indices (SI = log(IAP/Ksp)) calculated at 90 °C from the corrected free Fe³⁺ activity using the following solubility products (25 °C, dissolution with H⁺): goethite log K = −1.20, amorphous Fe(OH)₃ log K = 1.00, schwertmannite log K = −18.0, jarosite log K = −12.40 [21]; all constants were corrected to 90 °C by the van’t Hoff equation. The resulting saturation indices are reported in

Table 6. Saturation indices of potential Fe precipitates during cyclic leaching.

Stage	pH	TFe (g/L)	[Fe3+] Free (mol/L)	Goethite SI (90 °C)	Am. Fe(OH)3 SI (90 °C)	Schwertmannite SI (per-Fe, Free-Ion, 90 °C)	Schwertmannite SI (Total-Fe(III), 90 °C)
Primary	0.20	5.60	4 × 10−8	−7.5	−9.0	−5.1	−0.46
Cycle 1	0.45	9.69	6 × 10−8	−6.4	−7.9	−4.0	+0.64
Cycle 2	0.80	7.50	2 × 10−8	−6.0	−7.5	−3.7	+0.94
Cycle 3	1.05	5.00	1 × 10−8	−5.7	−7.2	−3.5	+1.14
Cycle 4	1.31	2.90	4 × 10−9	−5.4	−6.8	−3.2	+1.44

. Goethite (SI −7.5 to −5.4) and amorphous Fe(OH)₃ (SI −9.0 to −6.8) remain far undersaturated at all stages, ruling out conventional hydrolytic Fe³⁺ precipitation as the removal mechanism. Schwertmannite-type basic ferric sulfate is the phase whose saturation index is most sensitive to pH (per-Fe SI from −5.1 to −3.2 in free-ion terms). In terms of the total dissolved Fe(III)—the reservoir from which precipitation draws as sulfate complexes dissociate—the solution becomes increasingly supersaturated with respect to this phase from Cycle 2 onward (pH ≈ 0.80), coinciding precisely with the onset of net Fe removal (Section 3.5). Jarosite-type KFe₃(SO₄)₂(OH)₆ remains undersaturated in free-ion terms (SI ≈ −10 to −13 at [K⁺] ≈ 0.01 M) but may approach saturation if alkali cations accumulate in the recycled leachate [22,23].

The elevated temperature (90 °C) further suppresses goethite precipitation (dissolution endothermic, ΔH = +48.0 kJ/mol, 32-fold solubility increase vs. 25 °C) while accelerating schwertmannite formation kinetics [24,25]. Activity coefficients were calculated with the Davies equation (A = 0.5115 at 25 °C, 0.5972 at 90 °C); the estimated ionic strength (I about 1.2–1.7 mol/L) exceeds the nominal range of the Davies equation (I < 0.5 mol/L), and results are also reported at gamma = 1 to bracket the uncertainty; all equilibrium constants were corrected from 25 to 90 °C by the van’t Hoff equation (ΔH values: goethite +48, jarosite +20, Fe(III)-sulfate complexes −12 to −25, schwertmannite 0 kJ/mol). Saturation indices were computed in both free-ion and total-Fe(III) formulations. Nernst-estimated solution potentials (E(Fe³⁺/Fe²⁺) about 0.52 V at 90 °C) provide the redox trajectory in Figure 6. For the reduction of Fe³⁺ by residual sulfides (Reaction 6, Section 4.4), the cell potential (E(Fe³⁺/Fe²⁺) = +0.77 V vs. E(FeS₂/Fe²⁺, ${\mathrm{SO}}_4^{2-}$) about −0.4 to −0.35 V) gives ΔG about −1.5 × 10³ kJ/mol at 25 °C; the reaction remains thermodynamically favorable under the experimental conditions [26].

The Pourbaix (E–pH) diagram of the Fe–S–H₂O system at 90 °C (Figure 6) visualizes these thermodynamic relationships. The stability fields confirm that goethite (pH ≥ 2.32) and hematite (pH ≥ 1.86) lie far beyond the experimental window (pH ≤ 1.31). Schwertmannite-type basic ferric sulfate appears as a narrow metastable field at pH 1.60–1.69. Notably, the entire cyclic leaching trajectory remains within the Fe³⁺ (aq) field; the equilibrium diagram does not predict iron precipitation under these conditions. This gap between the equilibrium prediction and the measured 48.2% Fe removal is consistent with the formation of metastable schwertmannite-type basic ferric sulfate, which is kinetically favored at 90 °C and driven by total-Fe(III) supersaturation (Table 6). The trajectory potentials are Nernst-estimated from measured Fe³⁺/Fe²⁺ ratios (E° ≈ 0.70 V at 90 °C; Fe³⁺/Fe²⁺ boundary at E ≈ 0.52 V after activity correction) and labeled “Estimated from Fe³⁺/Fe²⁺. The Nernst-estimated potential decreases from about +0.75 V (Cycle 1) to +0.65 V (Cycle 4, SHE), reflecting the progressively reducing environment (Section 3.6).

4.2. Mineral Phase Evolution—XRD Analysis

To elucidate mineral phase evolution during cyclic leaching, XRD analysis was conducted on the PyC and the residues from each leaching cycle, as shown in Figure 7. All samples exhibited diffraction peaks corresponding to α-Fe₂O₃ (PDF 97-002-2505), with no measurable shifts in peak position, shape, or relative intensity [25]. The persistence of α-Fe₂O₃ as the dominant phase confirms that leaching proceeds via surface dissolution of Cu²⁺, Co²⁺, and Fe³⁺ ions, accompanied by solution-phase reactions, without altering the bulk crystal structure of the hematite matrix.

Notably, no diagnostic diffraction peaks for schwertmannite (2θ = 18.1°, 26.2°, 35.5°) or goethite (2θ = 17.8°, 21.2°, 36.6°) were observed in the residues. This is attributed to three factors: (i) the poor crystallinity and broad diffraction maxima characteristic of schwertmannite formed under rapid precipitation conditions, and (ii) the very low mass fraction of the precipitated Fe phases: the dissolved Fe removed between the Cycle-1 peak and Cycle 4 (ΔTFe ≈ 6.79 g/L) corresponds to ≈1.2 g of Fe (≈2 g of schwertmannite), i.e., <1 wt% of the residue, far below the detection limit of conventional powder XRD, particularly for a poorly crystalline phase with broad, low-intensity maxima; (iii) the high fluorescence background of the Fe-rich samples under Cu Kα radiation (≈50% of the main-peak intensity), which further obscures any weak, broad feature of a poorly crystalline minority phase.

It should be noted that adsorption and heterogeneous uptake of Fe³⁺ and sulfate on the surfaces of the freshly added hematite matrix are fully consistent with the surface-sensitive XPS results and with the unchanged bulk XRD patterns, and may contribute to the observed Fe removal. Three observations nonetheless support precipitation as the dominant pathway: (i) the onset of net Fe removal (Cycle 2) coincides with the cycle in which the solution first reaches supersaturation with respect to schwertmannite-type basic ferric sulfate in total-Fe(III) terms, whereas adsorption would not require supersaturation; (ii) the solid-phase sulfur accumulation continues monotonically across successive batches and is stoichiometrically consistent with the Fe removed (Fe/S ≈ 8:1, Section 4.3); and (iii) in the strongly acidic, sulfate-rich medium, sulfate inhibits the nucleation of goethite and hematite, whereas schwertmannite-type phases nucleate readily and precipitate within hours at 90 °C, consistent with the 2 h cycle time. A minor contribution of adsorption cannot be fully excluded and is acknowledged; direct phase identification of the precipitate is beyond the scope of the present work and is identified as a priority for future work.

4.3. XPS Analysis—Fe 2p and S 2p

To further elucidate the intrinsic mechanisms underlying Fe valence evolution and the nature of Fe removal, XPS was employed for peak fitting and semi-quantitative analysis of Fe and S species in PyC and leach residues obtained after successive cyclic leaching cycles (Figure 8;

Table 7. Relative content of iron valence states determined by XPS.

Sample	Fe3+ (at%)	Fe2+ (at%)
PyC	81.95	18.05
Leached Residue 0	87.99	12.01
Residue 1	88.22	11.78
Residue 2	86.14	13.86
Residue 3	84.42	15.58
Residue 4	83.46	16.54

).

The relative contents of Fe³⁺ and Fe²⁺ were calculated based on peak area ratios [27,28], as presented in Figure 8a and Table 7. The Fe²⁺ content in the residue initially decreased (PyC → Residue 1: 18.05% → 11.78%) and then progressively increased (Residue 1 → Residue 4: 11.78% → 16.54%). Simultaneously, the Fe²⁺ proportion in solution increased continuously throughout the process (reaching 82.80% of TFe by Cycle 4). These trends reveal a dynamic interplay between Fe dissolution, reduction, and precipitation.

4.3.1. Fe 2p Analysis

Stage 1 (Primary leaching): Strongly acidic conditions (pH 0.20) promoted rapid dissolution of native Fe²⁺ from partially roasted sulfides and low-valence Fe oxides. Owing to the higher acid-solubility of these Fe(II)-bearing phases relative to the hematite matrix, the solid residue is enriched in Fe(III), reflected in the decreasing Fe²⁺ proportion from 18.05% (PyC) to 11.78% (Residue 1, Table 7). The Fe³⁺ that subsequently participates in the reduction and precipitation steps originates from direct acid dissolution of Fe(III) oxides in the cinder, not from re-oxidation of dissolved Fe²⁺.

Stage 2 (Cycles 2–4): As the residual H₂SO₄ concentration declined and pH rose, partially roasted sulfides and low-valence metal oxides in freshly added PyC progressively dissolved, enhancing the reducing capacity of the aqueous phase. This redox shift drove two interconnected processes: (i) aqueous Fe³⁺ dissolved in earlier cycles was reduced to Fe²⁺; and (ii) a significant portion of the Fe³⁺ was removed from solution, consistent with schwertmannite-type basic ferric sulfate precipitation. The Fe²⁺ remained in solution as stable sulfate complexes and was not incorporated into the precipitating phase.

4.3.2. S 2p Analysis

The S 2p spectra were analyzed to track the sulfur speciation and relative content in the solids (Figure 8b, Table 5). The PyC initially contained the highest sulfur content (relative peak area: 19,695). Following primary leaching, the S content in the residue decreased sharply to 5375, directly attributable to the high initial acid concentration facilitating the dissolution of soluble sulfate species.

The S 2p spectra of all samples are dominated by a single broad peak centered at ≈168–170 eV, characteristic of sulfate-type sulfur (${\mathrm{SO}}_4^{2-}$). Minor intensity at lower binding energies (≈161–165 eV), attributable to residual sulfide and elemental sulfur, is present in all spectra but constitutes a small fraction of the total S 2p signal.

Critically, the relative total S 2p peak area increased monotonically across the cyclic leaching sequence: 11,072 (Residue 1) → 11,302 (Residue 2) → 12,664 (Residue 3) → 14,851 (Residue 4), representing a 176% increase relative to the Residue 0 minimum (5375). This progressive sulfur accumulation in the solid phase cannot be explained by the common ion effect, which would only suppress further dissolution of sulfur-bearing phases rather than introduce additional sulfur into the solid. Residual entrained leachate is also unlikely to account for the increase: constant liquid retention would produce a roughly constant—not monotonically increasing—S signal. The accumulation is therefore attributable to the progressive formation of a sulfur-bearing solid phase.

The S 2p sulfate signal alone cannot distinguish among candidate sulfate-bearing phases: adsorbed sulfate, jarosite-type phases, and schwertmannite-type phases all exhibit S 2p binding energies in the 168–170 eV range. However, the quantitative relationship between the Fe removed and the S accumulated provides a discriminating constraint. The dissolved Fe removed between the Cycle-1 peak and Cycle 4 (ΔTFe ≈ 6.79 g/L, corresponding to ≈1.2 g Fe over four cycles) and the associated sulfur accumulation yield a molar Fe/S ratio consistent with Fe₈O₈(OH)₆SO₄ stoichiometry (Fe/S = 8). Jarosite-type phases (KFe₃(SO₄)₂(OH)₆, Fe/S = 1.5) would require substantially more sulfur than is observed.

The dominance of the sulfate peak in every spectrum, together with the Fe/S stoichiometric consistency, indicates that the accumulating sulfur is predominantly sulfate and is best explained by incorporation into a sulfate-bearing Fe(III) precipitate. A minor contribution from incomplete sulfide oxidation to elemental sulfur (S⁰, which would appear at ≈163.5–164.5 eV) cannot be rigorously excluded; however, the position of the dominant S 2p peak at ≈168–170 eV in all samples is inconsistent with S⁰ being the primary accumulating species.

Combined with the thermodynamic analysis (Table 6: supersaturation with respect to schwertmannite-type basic ferric sulfate from Cycle 2 onward) and the Fe mass balance (Table 4), the converging evidence is best explained by schwertmannite-type basic ferric sulfate precipitation.

4.4. Corrected Reaction Network and Synergistic Mechanism

Based on the combined thermodynamic, XRD, and XPS evidence, we propose the following reaction network governing the coupled Fe removal and Cu-Co enrichment during neutralizer-free cyclic leaching of PyC.

• Step 1: Acid Dissolution and Residual Acid Consumption

Metal oxides in fresh PyC react with free H⁺ in the recycled leachate via acid dissolution:

Me_xO_y + 2yH⁺ → xMe^(2y/x)+ + yH₂O (Me = Cu, Co, Fe, Ca, Mg)

(4)

This process extracts Cu and Co into the aqueous phase while simultaneously consuming residual H₂SO₄. The primary H⁺ consumers are basic oxides (CuO, CoO, CaO) rather than Fe₂O₃, which is only sparingly soluble under the conditions of later cycles.

• Step 2: Fe³⁺-${\mathrm{SO}}_4^{2-}$ Complexation

In the high-sulfate environment ([${\mathrm{SO}}_4^{2-}$] = 1 mol/L), dissolved Fe³⁺ exists predominantly as sulfate complexes:

$${\mathrm{Fe}}^{3+}+\mathrm{n}{\mathrm{SO}}_4^{2-}\to {\mathrm{FeSO}}_4^{+}/{\mathrm{Fe}\left(\mathrm{S}{\mathrm{O}}_4\right)}_2^{-}$$

(5)

The strong complexation (α ≈ 251,000) suppresses the free Fe³⁺ concentration to ~10⁻⁸ mol/L, preventing simple hydrolytic precipitation as Fe(OH)₃ or goethite.

• Step 3: Endogenous Reduction of Fe³⁺

The Fe²⁺/TFe ratio increases across the cycles because each fresh batch of PyC supplies Fe²⁺ through acid dissolution of Fe²⁺-bearing phases—fayalite-type Fe₂SiO₄, magnesioferrite-type spinels (Table 2), and residual low-valence Fe species—while Fe³⁺ is selectively removed from solution by schwertmannite-type precipitation (Step 4). The Fe²⁺ content of the cinder (1.37 wt%, Table 1) is sufficient to account for the observed increase in solution Fe²⁺ (Section 3.6). Residual sulfur species in the cinder (0.13 wt% total S, Table 1) may contribute additional reducing equivalents through the thermodynamically favorable oxidation of reduced sulfur by Fe³⁺ [26] (Reaction 6).

$${\mathrm{FeS}}_2+14{\mathrm{Fe}}^{3+}+8{\mathrm{H}}_2\mathrm{O}\to 15{\mathrm{Fe}}^{2+}+2{\mathrm{SO}}_4^{2-}+16{\mathrm{H}}^{+}$$

(6)

Reaction 6 is thermodynamically favorable under the experimental conditions (ΔG < 0; Section 4.1), and the 16H⁺ generated per mole of FeS₂ partially offset the acid consumption of Step 1. The resulting Fe²⁺ forms stable sulfate complexes and remains in solution.

• Step 4: Schwertmannite Precipitation (Primary Fe Removal Mechanism)

The key iron removal mechanism is the precipitation of basic ferric sulfate (schwertmannite):

$$8{\mathrm{Fe}}^{3+}+{\mathrm{SO}}_4^{2-}+14{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Fe}}_8{\mathrm{O}}_8(\mathrm{OH}{)}_6{\mathrm{SO}}_4\downarrow +22{\mathrm{H}}^{+}$$

(7)

Schwertmannite-type basic ferric sulfate is the Fe phase whose precipitation is favored over goethite and Fe(OH)₃ in this acidic sulfate medium: its saturation index rises steeply with pH, and the solution enters the supersaturated regime (in total-Fe(III) terms) as the pH approaches ≈0.8 (Cycle 2; Table 6), while goethite and Fe(OH)₃ remain far below saturation throughout. The reaction is kinetically accelerated at 90 °C, with formation reported within hours under comparable conditions [29]. This precipitation pathway exhibits two critical characteristics:

It consumes Fe³⁺ directly, resulting in the observed decrease in dissolved Fe³⁺ from 8.2 g/L (Cycle 1) to 0.49 g/L (Cycle 4).

It is H⁺-producing (2.75 mol H⁺ per mol Fe), meaning the cyclic leaching system operates with competing acid-generating and acid-consuming reactions. Across all cycles, the acid consumed by oxide dissolution (Reaction 4) exceeds the acid generated by sulfide oxidation (Reaction 6) and schwertmannite precipitation (Reaction 7), consistent with the observed net pH increase; the high-sulfate medium buffers the pH response through the HSO₄-/${\mathrm{SO}}_4^{2-}$ equilibrium.

A secondary precipitation pathway may involve jarosite formation if sufficient alkali cations (K⁺, Na⁺) are present:

$$3{\mathrm{Fe}}^{3+}+2{\mathrm{SO}}_4^{2-}+{\mathrm{K}}^{+}+6{\mathrm{H}}_2\mathrm{O}\to \mathrm{K}{\mathrm{Fe}}_3\left({\mathrm{SO}}_4{)}_2\right(\mathrm{OH}{)}_6\downarrow + 6{\mathrm{H}}^{+}$$

(8)

• Step 5: Partial Conversion to Goethite (Minor Pathway)

At 90 °C, schwertmannite is metastable and may partially transform to goethite over time [22]:

$${\mathrm{Fe}}_8{\mathrm{O}}_8(\mathrm{OH}{)}_6{\mathrm{SO}}_4+2{\mathrm{H}}_2\mathrm{O}\to 8(\mathrm{\alpha }-\mathrm{FeOOH})+{\mathrm{SO}}_4^{2-}+2{\mathrm{H}}^{+}$$

(9)

Schwertmannite-type phases may undergo partial transformation to goethite at 90 °C. Based on reported transformation kinetics (half-life approximately 1–3 days at 80 °C), the extent of conversion during 8 h of cumulative leaching is expected to be limited (<20–50%), consistent with the absence of goethite diagnostic peaks in XRD.4.5. Comparison with Conventional Processes and Process Significance.

In conventional acid leaching systems, Fe removal relies almost exclusively on hydrolytic precipitation of Fe³⁺ as hydroxides or oxyhydroxides, requiring pH adjustment to >2.5 via neutralizer addition. This approach increases reagent costs, introduces impurity ions (Ca²⁺, Mg²⁺) that interfere with downstream solvent extraction, and causes 5–15% co-precipitation loss of dissolved Cu and Co.

The proposed neutralizer-free cyclic leaching process achieves simultaneous Cu-Co enrichment and 48.2% Fe removal (70.1% relative to peak Fe concentration) under strongly acidic conditions (pH < 1.5) through a fundamentally different mechanism: schwertmannite-type basic ferric sulfate precipitation, enabled by the unique combination of high sulfate concentration, elevated temperature (90 °C), and the progressive pH increase inherent to cyclic operation.

This mechanism resolves a long-standing challenge in PyC hydrometallurgy: Fe removal at low pH without neutralizer addition and without co-precipitation of valuable Cu and Co. Copper and cobalt ions remain in solution as stable aqueous complexes and do not incorporate into the schwertmannite-type or jarosite crystal lattices during precipitation, enabling simultaneous selective enrichment of target metals in the liquid phase. This statement is supported by the quantitative residue analyses reported in Table 4: the Cu and Co contents of all residues (Cycles 1–4) are small, and the complete Cu/Co balance shows that the fraction of dissolved Cu and Co retained in the solids is dominated by un-leached metal from the fresh cinder rather than by co-precipitation with the Fe precipitate.

Comparison with single-cycle leaching. Although a dedicated single-cycle leaching baseline under identical conditions was not conducted in this experimental series, the primary leaching step (100 g/L H₂SO₄, L/S = 3:1, 90 °C, 2 h), which represents the initial state of the cyclic process, can serve as a meaningful baseline. The Cu concentration after primary leaching (1.98 g/L) is magnified to 9.42 g/L after 4 cycles (×4.76 enrichment), while the pH rises from 0.20 to 1.31. In a conventional batch process, achieving this pH adjustment would consume approximately 12 kg CaCO₃ per m³ of leachate, with an additional 5–15% loss of dissolved Cu and Co through co-precipitation with iron hydroxides. The cyclic process thus delivers simultaneous enrichment and acid consumption in a single operation.

Preliminary economic assessment. A simplified cost comparison further illustrates the process benefits. For a plant processing 100 m³/day of Cu-Co leachate, the conventional neutralization step consumes approximately 1.2 tonnes/day of CaCO₃, generating ~2.4 tonnes/day of gypsum residue. The cyclic process eliminates this reagent cost and waste stream entirely. Additionally, by avoiding Cu/Co co-precipitation losses (estimated 5–15% in conventional Fe(OH)₃ precipitation), the cyclic process recovers an additional 0.3–0.9 kg Cu and 0.06–0.18 kg Co per tonne of PyC processed, representing meaningful economic value at current metal prices without additional reagent expenditure.

5. Conclusions

A neutralizer-free cyclic leaching process was developed for Cu-Co-bearing pyrite cinder, simultaneously consuming residual acid, enriching Cu and Co, and removing Fe under strongly acidic conditions.

(1)

After 4 cycles, the pH rises from 0.20 to 1.31 with 96% residual H₂SO₄ consumed. Cu is enriched 4.76-fold to 9.42 g/L and Co 3.88-fold to 1.55 g/L, reaching concentrations suitable for downstream solvent extraction without neutralization.

(2)

An Fe removal of 48.2% is achieved (70.1% relative to the Cycle-1 peak), with Fe³⁺ decreasing from 8.2 to 0.49 g/L. The removal is best explained by schwertmannite-type basic ferric sulfate (Fe₈O₈(OH)₆SO₄) precipitation, consistent with thermodynamic saturation analysis (Table 6), the Fe mass balance (Table 4), and progressive solid-phase sulfate accumulation (XPS S 2p).

(3)

The mechanism involves four coupled processes. Acid dissolution of metal oxides consumes residual H₂SO₄, driving the pH increase. Strong Fe³⁺-${\mathrm{SO}}_4^{2-}$ complexation suppresses free Fe³⁺, preventing conventional hydrolytic precipitation. Fe²⁺ supplied by dissolution of Fe²⁺-bearing phases (fayalite, magnesioferrite-type spinels) in each fresh batch of cinder, combined with selective Fe³⁺ removal by precipitation, drives the Fe²⁺/TFe ratio from 15% (Cycle 1) to 83% (Cycle 4). Schwertmannite-type basic ferric sulfate precipitation becomes operative from Cycle 2 onward (pH about 0.80).

(4)

Unlike conventional hydrolytic Fe removal, this process requires no neutralizer, introduces no Ca²⁺/Mg²⁺ impurities, and avoids Cu/Co co-precipitation losses. It offers a sustainable strategy for valorizing pyrite cinder and other iron-rich secondary resources.