Chromium Control in Ammonium Vanadate Precipitation from Slag-Derived Liquors: A Diagnostic Framework Integrating Speciation, Entrainment, and Mother-Liquor Recycling

Abstract

Ammonium precipitation is widely used to recover vanadium from slag-derived liquors as ammonium metavanadate or polyvanadate, but chromium control is often evaluated less systematically than vanadium recovery. This focused process-diagnostic review draws on the topic-guided literature identified from academic literature sources, preliminary literature-discovery tools, and reference-list tracing. It presents a diagnostic framework integrating V/Cr speciation, cake entrainment, washing response, soluble-ion tracers (Na⁺, SO₄²⁻, Cl⁻), and recycle-accumulation assessment to identify whether product chromium originates from retained mother liquor, solid-associated transfer, or recycle accumulation. The analysis shows that high single-pass vanadium precipitation efficiency alone is insufficient; credible process assessment requires product-Cr source diagnosis, Cr(VI) control, purge/salt-recovery planning, and multi-cycle validation.

1. Introduction

Vanadium slag and related iron- and steel-metallurgical by-products are important secondary resources for vanadium recovery due to their enrichment in vanadium and compatibility with existing metallurgical processing circuits [1,2]. However, vanadium in these residues is frequently accompanied by chromium. Depending on the upstream roasting, leaching, and conditioning routes, chromium may remain in the solid residue, enter the solution as Cr(III), or be oxidized to soluble Cr(VI) [3,4]. This multi-phase distribution poses significant challenges for product purity, Cr(VI)-related occupational and environmental risks, discharge compliance, and residue stabilization.

Ammonium-assisted precipitation is widely utilized because V(V) can be selectively recovered as ammonium metavanadate (AMV) or ammonium polyvanadate (APV) through appropriate pH and salinity adjustments [1]. Ammonium-rich circuits typically originate from ammonium-sulfate leaching, ammonium conditioning of sodium vanadate liquors, or ammonium-assisted precipitation following solvent extraction [5,6,7]. Since the roasting and leaching routes dictate both the Cr(VI)/Cr(III) distribution and the Na⁺/NH₄⁺/SO₄²⁻/Cl⁻ salt matrix entering the precipitation stage, chromium control during ammonium vanadate precipitation must be assessed in conjunction with upstream process history [8].

Cr(VI) oxyanions generally remain soluble under weakly acidic-to-alkaline conditions, although their solubility in concentrated industrial liquors depends strongly on counterions, ionic strength and solution composition [9,10]. This behavior renders a vanadium-first (V-first) recovery strategy feasible: vanadium is selectively precipitated as AMV or APV while Cr(VI) remains predominantly in the mother liquor [5,9,11]. To close this metallurgical loop, subsequent processes should address this residual Cr(VI) in a downstream stage; for example, Yang et al. [12] developed a cyclic lead–salt precipitation–leaching route for chromium separation and recovery. However, dissolved chromium can still enter the vanadium product via mother-liquor entrainment or surface adsorption. In contrast, Cr(III) can hydrolyze, form colloids, co-precipitate with impurity metal hydroxides, or become physically trapped within vanadium filter cakes [13,14,15].

Although extensive work has addressed vanadium extraction and ammonium vanadate crystallization [9,16,17], many reported studies summarized in Tables S1–S4 still evaluate chromium control mainly using single-pass indicators, such as vanadium precipitation efficiency, total chromium removal, residual Cr concentration, or final product grade, rather than Cr contamination mechanism. These indicators fail to identify whether product chromium originates from retained mother liquor, surface adsorption, lattice occlusion, co-precipitation, or colloid capture. Furthermore, they do not illustrate the accumulation dynamics of Cr(VI) and salts during mother-liquor recycling. This review addresses three critical diagnostic gaps: chromium speciation, product-Cr uptake mechanism, and recycle stability.

The expressions employed are elementary mass balances, and no new mathematical formulation is claimed; the contribution lies in their integrated deployment as a sequential diagnostic procedure for identifying the origin of chromium in the vanadium product. Specifically, this review contributes to the literature by: (1) Establishing a speciation-based interpretation of chromium behavior during ammonium vanadate precipitation to distinguish soluble Cr(VI), hydrolyzed Cr(III), adsorbed chromium, and colloid-associated chromium. (2) Assembling an entrainment-based diagnostic procedure that bounds and screens measured product chromium against the theoretical chromium mass contributed by retained mother liquor. (3) Proposing the use of soluble-ion tracers such as Na⁺, SO₄²⁻ and Cl⁻ to differentiate between mother-liquor carryover and solid-associated chromium transfer. (4) Connecting product-Cr diagnosis with mother-liquor recycle inventory, thereby highlighting Cr(VI) and accumulation of dissolved salts (Na⁺, SO₄²⁻ and Cl⁻) as critical process-stability risks that cannot be evaluated by single-pass vanadium precipitation efficiency alone.

2. Scope, Evidence Base and Stream Definitions

In this review, “ammonium-rich” refers to precipitation or post-precipitation liquors in which NH₄⁺ is deliberately introduced as a major conditioning or precipitation reagent (e.g., via ammonium sulfate, ammonium chloride, ammonium bicarbonate, or ammonia-based pH adjustment), rather than being present only as a trace background ion.

2.1. Process Context

The circuits covered here follow a common sequence: oxidative roasting of vanadium slag (most commonly with sodium salts; routes are compared in Section 6.3), water or ammonium-sulfate leaching to give an alkaline, oxidizing liquor (pH ≈ 10.5–13 [5,8])—acid leaching (standard for calcification-roasted slag) converges on the same downstream steps—optional purification and solid–liquid separation, and ammonium-salt conditioning with pH adjustment to precipitate AMV or APV, followed by filtration and washing (Figure 1). Because oxidative roasting fixes vanadium as V(V), no separate Eh adjustment is needed upstream of precipitation; the same oxidizing conditions retain chromium as soluble Cr(VI).

Figure 1. Chromium-transfer pathways during ammonium vanadate precipitation and solid–liquid separation—the downstream steps of the sequence in Section 2.1. During AMV/APV precipitation, soluble Cr(VI)-bearing species may remain in the liquor, while Cr(VI) oxyanions may also adsorb onto AMV/APV crystals, and Cr(III) may associate with Fe/Al- or Si-bearing colloids/flocs. During filtration and washing, Cr can enter products through retained pore liquor, adsorption, co-precipitation or colloid capture.

2.2. Scope and Literature Selection

This work presents a focused narrative and process-diagnostic review rather than a systematic review or meta-analysis. The evidence base was drawn primarily from peer-reviewed journal articles and review papers. The literature was identified through searches in Google Scholar up to January 2026, with preliminary literature discovery assisted by AI-based semantic search as described in Section 2.4. No fixed starting year was imposed on the search. Searches used combinations of keywords including “vanadium slag” (a feedstock in which chromium is most significantly co-enriched with vanadium and co-extracted during roasting and leaching (Section 6.3)), “vanadium extraction”, “ammonium vanadate”, “ammonium metavanadate”, “chromium”, “Cr(VI)”, “chromate”, “precipitation”, “crystallization”, “separation”, “mother liquor”, “entrainment”, “washing”, and “salt recovery”, tailored to four thematic blocks: (i) vanadium slag processing with chromium co-extraction; (ii) ammonium vanadate precipitation and crystallization; (iii) V/Cr separation from post-precipitation liquor; and (iv) Cr speciation and reoxidation thermodynamics. Vanadium sources without significant chromium co-enrichment (e.g., direct processing of VTM ores or stone coal) were not systematically targeted. Retrieved records were screened by title and abstract. Publications were included when they provided data, evidence, or transferable methodology on at least one of the following aspects: chromium speciation, ammonium vanadate precipitation, impurity transfer to products, washing or filtration, residual V/Cr recovery, salt crystallization, or mother-liquor recycling. Records were excluded when chromium was not valence-resolved or when the precipitation stage was insufficiently described. No publication-year restriction was applied. Backward tracing of reference lists and forward citation tracing were used extensively to ensure coverage. The resulting evidence base comprises 56 cited publications, of which 32 primary studies meeting the chromium-relevance criteria are tabulated in Tables S1–S4

2.3. Process Streams and Chromium-Control Functions

The main process streams considered in this review are summarized in Table 1. They are distinguished because chromium-control functions and challenges differ throughout the circuit.

Table 1. Stream definitions and chromium-control issues in ammonium-rich vanadium precipitation circuits.

Stream	Definition	Main Chromium-Control Issue
Ammonium-rich feed liquor	V–Cr-bearing liquor before precipitation	Cr valence, V supersaturation, salt matrix
Precipitation slurry	Suspension during AMV/APV formation	Cr adsorption, hydrolysis, impurity capture
Vanadium filter cake	Solid vanadium product before/after washing	Mother-liquor entrainment, product Cr
Post-precipitation mother liquor	Filtrate after precipitation	Residual V, Cr(VI), total Cr, NH4+, Na+, sulfate
Recycled liquor	Liquor returned to the circuit	Salt accumulation, Cr(VI) build-up
Bleed/purge stream	Stream withdrawn from recycle loop	Residual V recovery, salt discharge, Cr(VI) compliance, trace-impurity removal

2.4. Evidence Limitations

The available literature is uneven in both scope and reporting detail. Most studies were designed to optimize vanadium recovery or product grade rather than to resolve chromium-transfer pathways or recycle evolution. Consequently, complete datasets linking liquor chemistry, solid–liquid separation, product impurities, and recycle behavior are uncommon. The proposed framework provides a diagnostic structure for hypothesis generation and consistency checking. A detailed evidence matrix summarizing key variables and reporting gaps across selected studies is provided in Tables S1–S4 in the Supplementary Materials.

2.5. Use of AI-Assisted Tools

AI-assisted tools were utilized during the preparation of this manuscript to improve preliminary literature identification and writing efficiency. Specifically, the semantic search engine Consensus (Pro subscription) was used as a supplementary tool to assist in the initial discovery of potentially relevant publications, after which formal literature searches were conducted in Google Scholar. All retrieved publications, bibliographic information, and technical data were manually verified and checked against the original sources by the authors. Additionally, GPT-based large language models (GPT-4/GPT-5 series) were employed to assist with research structuring, outline development, and language polishing. All retrieved studies, interpretations, and scientific arguments were critically reviewed and verified manually by the authors. No AI tools were used to generate raw research data, perform independent calculations, or draw independent scientific conclusions.

3. Speciation and Crystallization Basis of Chromium Control

3.1. V(V) Speciation and AMV/APV Crystallization

In ammonium-rich precipitation systems, vanadium recovery is governed by V(V) oxyanion protonation, condensation, and crystallization under specific pH, ionic-strength, and temperature conditions. In alkaline and weakly alkaline liquors, V(V) occurs mainly as monomeric or protonated vanadate species, while acidification and concentration promote polyvanadate formation, including decavanadate-like species [18,19]. These transformations determine vanadium activity and supersaturation; as vanadium is removed during precipitation, the equilibrium speciation shifts toward monomeric vanadates, leaving an increasingly dilute residual liquor whose speciation governs residual-vanadium recovery and chromium behavior in the recycled stream.

Ammonium vanadate precipitation is a crystallization–purification operation whose product phase, crystal habit, and filterability depend on pH trajectory, NH₄⁺ dosage, temperature, vanadium concentration, and co-existing impurities. Phase assignment (AMV vs. APV) should be supported by X-ray diffraction (XRD) or thermal analysis, which identify the bulk crystalline phase but do not resolve where impurities reside. Jiang et al. (2023) [17] showed that hydrolysis during acidic ammonium salt precipitation from high-concentration alkaline vanadium solutions can increase Na impurity in APV. In V–Cr systems, weakly acidic precipitation has been applied; for example, APV precipitation at pH 4.5, 293 K and 60 min gave a reported V precipitation efficiency of 94.8%, while Cr remained largely in solution, Cr in the product was not reported [5]. However, this pH is not a universal optimum because the precipitation window shifts with salt matrix, V/Cr ratio and recycle history.

Furthermore, dissolved chromium and co-existing salts can modify ammonium vanadate crystallization. Cr(VI) can increase the apparent solubility of NH₄VO₃, reduce supersaturation, alter crystal morphology, and raise Cr impurities in the final product; in NH₄VO₃ precipitation, increasing Cr(VI) from 2.0 to 25 g L⁻¹ decreased V precipitation efficiency from 83.5% to 60.3% [9]. Salt background matters as well: NH₄VO₃ solubility decreases markedly at high (NH₄)₂SO₄ and NH₄Cl concentrations [20], raising supersaturation; the resulting preference for nucleation over growth is expected to yield finer crystals with greater surface area and higher mother-liquor retention, as reflected in the entrainment of mother liquor by explosively formed fine crystals reported by Chen et al. [16]. Zhan et al. [21] further showed that impurities such as Na⁺, K⁺, Cr(VI), and H₂PO₄⁻ affect the nucleation kinetics of APV, with phosphate notably altering the crystal morphology from flake to spheroidal. Crystallization control offers a countermeasure: in the heterogeneous self-assembly process of Chen et al. [16], slow NH₃ supply via (NH₄)₂CO₃ pyrolysis confined nucleation to the container wall at low supersaturation; at a feed Cr(VI) of only 0.05 g L⁻¹ (pH 12.5, 60 °C), this route yielded 1.8 mm NH₄VO₃ crystals at 93.09% yield with only 1 ppm Cr, versus ~110 ppm for direct precipitation. Explicit seeding, though seldom reported for vanadate systems, could in principle act through the same surface-mediated mechanism in conventional stirred crystallizers.

3.2. Chromium Valence, Hydrolysis, and Residue Stability

Chromium behavior is governed primarily by the Cr(VI)/Cr(III) redox couple and by Cr(III) hydrolysis. Cr(VI) predominantly exists as chromate (CrO₄²⁻), hydrogen chromate (HCrO₄⁻), or dichromate-related (Cr₂O₇²⁻) oxyanions depending on pH [8]. These oxyanions are generally more soluble than Cr(III) hydroxide phases under the weakly acidic-to-alkaline conditions relevant to most vanadium precipitation circuits. Therefore, Cr(VI) remains in solution during V-first ammonium vanadate precipitation, shifting the primary chromium-control burden to the mother-liquor circuit. To mitigate this, various post-precipitation techniques, such as lead carbonate precipitation, have been developed to scavenge and recover residual chromium directly from these vanadium-precipitated solutions. In the cyclic route of Yang et al. [12], PbCO₃ precipitation removed chromium to 0.001 g L⁻¹ as PbCrO₄ while the dissolved vanadium remained unchanged (0.01 g L⁻¹), i.e., vanadium co-precipitation was negligible; the lead precipitant was regenerated by Na₂CO₃ leaching and the chromium-depleted liquor recycled to the vanadium circuit.

Cr(III) is readily hydrolyzed and precipitated. At pH ≈ 6–10, Cr(III) hydrolyzes and precipitates as Cr(OH)₃ or Cr-bearing hydroxide phases with minimum solubility around pH 8–9 [22]; above pH ≈ 11, amphoteric redissolution as Cr(OH)₄⁻ becomes appreciable. This hydrolysis behavior has been utilized in various separation processes [8,11,13] and can be simulated using thermodynamic software such as PHREEQC (version 3) [23]. However, if reduction or pH adjustment occurs before or during ammonium vanadate crystallization, Cr(III)-bearing hydroxides may form colloids or impurity-associated solids. These can impair filtration, increase cake moisture, trap mother liquor and promote chromium transfer into AMV/APV products.

Because this precipitation is pH-reversible, a decrease in total Cr in the separated liquor does not by itself demonstrate durable removal: the precipitate can redissolve at high pH, and colloidal Cr(III) may enter the analyzed fraction depending on sample filtration (Section 3.3 and Section 4.2). Although E–pH diagrams and speciation calculations provide theoretical thermodynamic limits, they must be validated with high-salt liquor analyses because highly concentrated electrolytes induce significant activity-coefficient deviations [8,10,14,23,24]. This reversibility likewise underlies the residue-stability evaluation below.

Cr(III)-bearing residues may be reoxidized under alkaline, oxic or Mn-bearing conditions [25,26]. Cr(VI) reduction should therefore be evaluated together with immediate removal efficiency and long-term residue stability.

3.3. High-Salt Matrix, Impurities and Analytical Considerations

The sulfate–sodium salt matrix itself exerts a strong ionic-strength effect on ammonium vanadate crystallization. High concentrations of (NH₄)₂SO₄ and NH₄Cl sharply decrease the solubility of NH₄VO₃ [20], so that for a given vanadium concentration the solution becomes more supersaturated; under such conditions, nucleation dominates over crystal growth, producing finer crystals that retain more mother liquor per unit mass of cake during filtration [15]. This ionic-strength effect is particularly relevant in sulfate-rich ammonium vanadate circuits where (NH₄)₂SO₄ concentrations can reach several mol·kg⁻¹ [20].

The effects of secondary impurities are condition-dependent. Slag-derived leach liquors are typically alkaline and oxidizing, with pH commonly between about 10.5 and 13 [5,8]. Full liquor compositions and precipitation conditions for these systems are compiled in Table S2 (Supplementary Materials). Silicate is most problematic when pH adjustment (particularly into the near-neutral range, pH ~6–9) or aging promotes polymerization of supersaturated silica, especially with Ca²⁺/Mg²⁺ acting as cross-linkers; for example, roasted vanadium slag leachate may contain g L⁻¹-level residual Si and require desiliconization before ammonium vanadate precipitation [27], while thiourea-dioxide reduction of Cr(VI) from industrial sodium vanadate liquor (pH 11.0, 90 °C) also removed ~60% of Si, because CO₂ released by thermal decomposition of the urea by-product lowers the pH and precipitates dissolved silicate as H₂SiO₃ together with the Cr(OH)₃ residue [8]. Fe(III) and Al(III) hydrolyze at pH ≈ 3–4 and ≈4.5–5, respectively, and the points of zero charge of ferrihydrite and Al(OH)₃ lie near pH 7–8 [28,29] and 8–9 [30], respectively; these solids therefore remain positively charged throughout the weakly acidic-to-neutral precipitation window (e.g., pH 4.5–8) and adsorb chromate and vanadate oxyanions via electrostatic attraction. Consequently, Fe/Al carried into the precipitation stage transfers chromium from solution to the solid phase rather than merely consuming the reagent. Similarly, Cr(III) hydrolysis products (e.g., Cr(OH)₃), which form over the pH range given in Section 3.2, can co-precipitate with or become entrained in the vanadium product if they are generated during ammonium vanadate precipitation. In sulfate-rich ammonium vanadate liquors, gypsum (CaSO₄·2H₂O) becomes supersaturated once the ion activity product [Ca²⁺][SO₄²⁻] exceeds its solubility product (K_sp ≈ 2.4–2.5 × 10⁻⁵ mol² L⁻² at 25–40 °C) [31], meaning that even tens of mg L⁻¹ Ca can trigger scaling when sulfate is in the tens of g L⁻¹ range. Mg(OH)₂ precipitation is significant only above pH ≈ 9.5 under typical liquor Mg concentrations [23]. Accordingly, Ca/Mg scaling should be evaluated using saturation indices of gypsum, carbonate or hydroxide phases under the measured pH, temperature, sulfate/carbonate content and ionic strength, rather than from total Ca/Mg concentration alone [23].

Reliable chromium control requires valence-specific analysis using standard methods. Cr(VI) can be measured by 1,5-diphenylcarbazide UV–vis or ion chromatography, while Cr(III) is often estimated by difference between total Cr and Cr(VI) [32,33,34]. In vanadium slag leach liquors, high V and Fe/Al concentrations can interfere with Cr(VI) determination; colloidal particles not only adsorb dissolved Cr(VI) (biasing the dissolved fraction to be low) but their turbidity also scatters/absorbs light in colorimetric reading (biasing it to be high) if samples are not filtered appropriately; and redox-sensitive samples require prompt analysis or preservation to avoid valence changes during storage. Samples should be reported with filtration method, membrane pore size, dilution factor, preservation, storage time, pH, temperature and Cr(VI) method.

pH and oxidation–reduction potential (ORP) are useful operational indicators but not thermodynamic descriptors. Measured ORP is governed mainly by the high ionic strength (via activity coefficients) and by the dominant redox couples (V(V)/V(IV), Cr(VI)/Cr(III), and Fe(III)/Fe(II)) that set the mixed potential, rather than by the reference electrode or junction potential. E–pH diagrams and speciation models can delimit feasible redox and precipitation windows, but these are theoretical starting points only and the actual window must be confirmed experimentally, since they cannot represent activity effects, metastable phases, nucleation, aging, adsorption, colloid formation or mother-liquor entrainment in real high-salt liquors [10,23,24].

4. Chromium Transfer into Ammonium Vanadate Products

Chromium can contaminate ammonium vanadate products through three main pathways: mother-liquor entrainment, surface adsorption, and co-precipitation or colloid capture. These pathways are controlled by chromium valence, AMV/APV crystal size and morphology, filtration behavior, washing efficiency and impurity hydrolysis. They are illustrated schematically in Figure 1.

The entrainment and cake-washing equations used in this section are conventional mass-balance tools. Their value here is diagnostic rather than mathematical: when combined with Cr valence analysis, mother-liquor chemistry, cake moisture, washing response and soluble-ion tracers such as Na⁺, SO₄²⁻ and Cl⁻, they can test whether the measured product Cr is quantitatively consistent with retained Cr-bearing mother liquor. If the calculated entrainment contribution is much lower than the measured product Cr, solid-associated pathways such as adsorption, occlusion, co-precipitation or colloid capture should be investigated.

4.1. Mother-Liquor Entrainment: Diagnostic Balance and Washing Response

Mother-liquor entrainment is a major but often under-quantified pathway. When chromium remains mainly in the liquid phase, product contamination may still occur through retention of Cr-bearing mother liquor in filter-cake pores. Thus, favorable chemical V/Cr separation does not guarantee low chromium in the dried or calcined product.

Unless otherwise stated, liquid concentrations $C_i$ are expressed in mg L⁻¹, liquid volumes $V_{\mathrm{ML}}$ in L, dry product mass $m_{\mathrm{dry}}$ in kg, product impurity contents $w_{i,\mathrm{prod}}$ in mg kg⁻¹ on a dry AMV/APV basis, and ${\mathsf{\rho }}_{\mathrm{ML}}$ in kg L⁻¹. Cake moisture $\mathsf{\theta }$ is defined on a wet-mass basis. Solid and solution concentrations must use the same chemical basis, e.g., Cr as Cr, sulfur as S or sulfate, and chlorine as Cl or Cl⁻. If products are calcined to V₂O₅, impurity contents should be converted using the appropriate mass yield and reporting basis.

For diagnostic accounting, the measured product Cr can be conceptually partitioned as:

$$m_{\mathrm{Cr},\mathrm{prod}}{= m}_{\mathrm{Cr},\mathrm{ent}} + m_{\mathrm{Cr},\mathrm{ads}} +{ m}_{\mathrm{Cr},\mathrm{cp}}$$

(1)

where $m_{\mathrm{Cr},\mathrm{prod}}$ is the total chromium mass in the dried product, $m_{\mathrm{Cr},\mathrm{ent}}$ is the chromium mass attributed to mother-liquor entrainment, $m_{\mathrm{Cr},\mathrm{ads}}$ is adsorbed chromium, and $m_{\mathrm{Cr},\mathrm{cp}}$ is chromium associated with co-precipitation or colloid capture. It must be emphasized that none of the three terms is directly measurable: m_Cr,ent is inferred from the retained mother-liquor volume, itself obtained from tracer-based estimates subject to the uncertainties detailed below, while m_Cr,ads and m_Cr,cp both reside in the solid phase and cannot be separated by a solution mass balance. Equation (1) therefore serves as a conceptual screening framework rather than a rigorous mechanistic partitioning tool. For practical purposes, the two solid-associated terms are treated as a single lumped quantity:

$$m_{\mathrm{Cr},\mathrm{solid}}{= m}_{\mathrm{Cr},\mathrm{ads}} +{ m}_{\mathrm{Cr},\mathrm{cp}}$$

(2)

This lumped term is obtained by the difference from the overall balance:

$$m_{\mathrm{Cr},\mathrm{solid}}{= m}_{\mathrm{Cr},\mathrm{prod}} -m_{\mathrm{C}\mathrm{r},\mathrm{ent}}$$

(3)

It thus carries the combined uncertainty of its input quantities and should, whenever possible, be reported as a range with propagated uncertainty rather than as a single value. Further distinguishing the individual contributions of adsorption, co-precipitation, and colloid capture to m_Cr,solid requires independent solid-phase characterization (e.g., sequential extraction, XPS, or SEM–EDS) and is beyond the scope of the present balance. In screening applications, if m_Cr,ent is close to m_Cr,prod and washing removes Cr in parallel with conservative tracers, retained mother liquor is sufficient to explain most product chromium; if m_Cr,ent is substantially lower, solid-associated pathways (m_Cr,solid) become the dominant contribution and warrant further investigation.

The entrainment contribution can be estimated as:

$$m_{\mathrm{Cr},\mathrm{ent}} ={ C}_{\mathrm{Cr},\mathrm{ML}}{V}_{\mathrm{ML},\mathrm{res}}$$

(4)

where $C_{\mathrm{Cr},\mathrm{ML}}$ is the chromium concentration in the mother liquor, and $V_{\mathrm{ML},\mathrm{res}}$ is the equivalent volume of original mother liquor retained after filtration and washing.

If physical entrainment dominates, the entrainment-derived chromium content in the dry product is:

$$w_{\mathrm{Cr},\mathrm{ent}} = {\displaystyle \frac{C_{\mathrm{Cr},\mathrm{ML}}{V}_{\mathrm{ML},\mathrm{res}}}{m_{\mathrm{dry}}}}$$

(5)

where $m_{\mathrm{dry}}$ is the dry mass of ammonium vanadate product. If $C_{\mathrm{Cr},\mathrm{ML}}$ is expressed in mg L⁻¹, $V_{\mathrm{ML},\mathrm{res}}$ in L and $m_{\mathrm{dry}}$ in kg, then $w_{\mathrm{Cr},\mathrm{ent}}$ is obtained in mg kg⁻¹.

For an unwashed cake, the initial retained mother-liquor volume is:

$$V_{\mathrm{ML},0} = {\displaystyle \frac{m_{\mathrm{dry}}\mathsf{\theta }}{(1-\mathsf{\theta }){\mathsf{\rho }}_{\mathrm{ML}}}}$$

(6)

where $\mathsf{\theta }$ is the wet-basis cake moisture and ρ_ML is the density of the mother liquor. After washing:

$$V_{\mathrm{ML},\mathrm{res}} = V_{\mathrm{ML},0}{R}_{\mathrm{ML},\mathrm{W}}$$

(7)

where $R_{\mathrm{ML},\mathrm{W}}$ is the residual fraction of original mother liquor remaining in the cake at wash ratio W, defined as the ratio of the cumulative volume of washwater applied to the volume of mother liquor initially retained in the unwashed cake. Thus:

$$w_{\mathrm{Cr},\mathrm{ent},\mathrm{W}} = {\displaystyle \frac{C_{\mathrm{Cr},\mathrm{ML}}{V}_{\mathrm{ML},0}{R}_{\mathrm{ML},\mathrm{W}}}{m_{\mathrm{dry}}}}$$

(8)

$R_{\mathrm{ML},\mathrm{W}}$ is process-dependent and should be determined experimentally via washing curves or tracer tests [15,35,36], a practical tracer mass-balance estimator is given in Equation (10) below. In ammonium vanadate systems, washing should also be evaluated for vanadium loss, which depends on the wash-liquor composition: pure water may partially dissolve the precipitate [17], whereas ammonium-salt-bearing liquors suppress dissolution through the common-ion effect—for AMV, (NH₄)₂SO₄ or NH₄Cl additions lower NH₄VO₃ solubility by up to two orders of magnitude [20]. This risk is greater for AMV than for the less-soluble APV [37].

For illustrative calculations, wet-basis cake moisture of 0.25–0.35 and a mother-liquor density around 1.15–1.25 kg L⁻¹ (the latter corresponding to a sulfate/ammonium-rich solution) are assumed as representative ranges; these values are illustrative estimates rather than literature-validated upper limits. The following calculations use θ = 0.30 and ρ_ML = 1.20 kg L⁻¹ as representative values. These θ and ρ_ML values are illustrative assumptions, as they are not reported in the cited sources (Section 6.4); the corresponding liquor compositions and precipitation conditions are given in Table S2.

An apparent entrainment contribution ratio is:

$$F_{\mathrm{ent},\mathrm{W}} = {\displaystyle \frac{w_{\mathrm{Cr},\mathrm{ent},\mathrm{W}}}{w_{\mathrm{Cr},\mathrm{prod}}}}$$

(9)

where $w_{\mathrm{Cr},\mathrm{prod}}$ is the measured total chromium content in the dried product. F_ent,W must be read as an upper limit on the entrainment contribution. F_ent,W < 1 places a lower limit of (1 − F_ent,W) on the solid-associated share, whereas F_ent,W ≈ 1 shows only that retained mother liquor is sufficient to account for the product chromium, because adsorbed Cr-bearing species with a washing response similar to that of entrained solute cannot be distinguished on this basis alone.

The most useful extension of the entrainment balance is the use of conservative or semi-conservative soluble impurities as internal tracers of retained mother liquor. In slag-derived ammonium-rich liquors, Na⁺ and Cl⁻ can serve this purpose; sulfate may also be used provided that its concentration remains below the saturation limit of sulfate salts under the operating conditions. However, the application of tracer components carries inherent inaccuracies. Ammonium vanadate precipitates, particularly when fine or needle-like, often exhibit large specific surface areas, and adsorption of tracer ions onto the solid surface cannot be neglected; this adsorption adds to w_j,prod independently of entrainment, making the tracer-based residual fraction defined below an upper limit of the residual mother-liquor fraction.

The adsorption propensity differs among tracers: Cl⁻ is generally the least strongly adsorbed onto oxide/hydroxide surfaces; Na⁺ may undergo adsorption or ion exchange, particularly on Cr(III)-, Fe- or Al-bearing hydroxide residues; and sulfate may both adsorb and precipitate (e.g., as gypsum) in Ca-bearing liquors. Two practical safeguards are recommended. First, at least two tracers with contrasting surface affinities (e.g., Cl⁻ and Na⁺) should be applied in parallel: mutually consistent R_j,W values strengthen the entrainment interpretation, whereas systematic divergence (e.g., R_Na,W consistently exceeding R_Cl,W) signals adsorption or precipitation artifacts, in which case the lower value is preferred. Second, the washing curve for each tracer should be examined: physically entrained solute is displaced rapidly and decreases approximately exponentially with wash ratio, whereas an adsorbed fraction desorbs slowly and produces asymptotic behavior at high wash ratios. Where feasible, the magnitude of surface adsorption can be constrained independently, e.g., by BET surface-area measurement combined with adsorption isotherms of the tracer ions on the product.

For a tracer component j, such as Na, Cl or SO₄²⁻, the apparent residual fraction of original mother liquor after washing can be estimated as:

$$R_{j,\mathrm{W}} = {\displaystyle \frac{w_{j,\mathrm{prod}}{m}_{\mathrm{dry}}}{C_{j,\mathrm{ML}}{V}_{\mathrm{ML},0}}}$$

(10)

where $w_{j,\mathrm{prod}}$ is the content of component $j$ in the dried product, $m_{\mathrm{dry}}$ is the dry product mass, and $C_{j,\mathrm{ML}}$ is the concentration of the same component (j) in the mother liquor before washing, and $V_{\mathrm{ML},0}$ is the initial volume of retained mother liquor in the unwashed cake. In this equation, $w_{j,\mathrm{prod}}$ is preferably expressed in mg kg⁻¹ on a dry-product basis, $m_{\mathrm{dry}}$ in kg, $C_{j,\mathrm{ML}}$ in mg L⁻¹, and ($V_{\mathrm{ML},0}$) in L; $R_{j,\mathrm{W}}$ is therefore dimensionless.

The analytical basis must be identical between solid and solution before applying Equation (10). For example, sulfur should be reported consistently either as elemental S or as sulfate; chromium as elemental Cr or as Cr₂O₃; chlorine as Cl or Cl⁻. If different bases are used, stoichiometric conversion is required. Even when incorporation or precipitation of tracer j into the crystal lattice is negligible, surface adsorption on the high-specific-area precipitate inflates w_j,prod and makes R_j,W an upper limit on the true residual mother-liquor fraction rather than a true value. If the washwater contains a non-negligible concentration of j, its contribution should be included in the tracer mass balance when calculating R_j,W. This may occur with the ammonium-salt washes, which introduce SO₄²⁻ or Cl⁻. Under these conditions, Na⁺ is the preferred tracer, provided that the washwater is sodium-free.

Given the indirect nature of both the entrainment and tracer estimates, the differentiation between entrainment-dominated and solid-associated Cr should be presented as upper and lower limits rather than exact fractions. If the apparent retention fractions for Cr and the soluble-ion tracers (Na⁺, Cl⁻, SO₄²⁻) are similar across multiple tracers and decrease in parallel during washing, mother-liquor entrainment is likely the dominant pathway for product chromium. Conversely, if Cr persists in the product after washing has effectively removed Na, Cl, and SO₄²⁻, or if R_j,W values from different tracers diverge substantially, then adsorption, occlusion, co-precipitation, or colloid capture should be considered. This interpretation should, where possible, be supported by independent solid-phase characterization, such as XPS for surface Cr speciation or validated sequential extraction for operationally defined Cr associations, rather than by the solution balance alone.

4.2. Adsorption, Co-Precipitation and Colloid Capture

Besides entrainment, chromium may transfer to vanadium solids by surface adsorption [9]. This pathway becomes more important when the specific surface area of the crystals is high, pH favors adsorption [9], or impurity hydroxides provide reactive surfaces. It should be stressed that a large apparent particle size does not necessarily imply a low surface area: larger particles formed by agglomeration of fine crystallites retain the high internal surface area of their constituents, which remain accessible to the mother liquor, and therefore do not suppress surface-related uptake of contaminants. A genuine reduction in surface-related Cr uptake requires the growth of truly larger single crystals, which in practice is achievable only by controlled crystallization with proper seeding, not by spontaneous precipitation [16]. Under such conditions, surface-related Cr uptake can be reduced and product purity improved. In contrast, Cr(VI) had little effect on APV nucleation [21].

Co-precipitation and colloid capture are especially relevant when chromium exists as Cr(III). Cr(OH)₃, Fe(OH)₃, Al(OH)₃ and silica gels can impair filtration, trap mother liquor and attach to vanadium product crystals. High-Al and sulfate-bearing vanadium systems also show that Al/Na/SO₄²⁻ precipitation chemistry can affect impurity behavior [38]. Therefore, product Cr should not be interpreted solely as a V/Cr separation problem.

Cr(III)-bearing colloids should be distinguished operationally from truly dissolved Cr(III) using different pore-size filtration (e.g., unfiltered, 0.45 μm, and finer filters). The difference between unfiltered total Cr and 0.45 μm filtered Cr represents coarse suspended or particulate Cr; the difference between 0.45 μm filtered and finer-filtered Cr provides an operational estimate of colloidal Cr. Acidification before filtration should be avoided when colloid speciation is the target.

No single solid-characterization technique resolves adsorbed chromium, impurity-phase chromium and physically entrained mother liquor; XRD in particular identifies only the major crystalline phases and gives no information on the composition of individual phases or on adsorbed ions. Surface-analytical methods (XPS), XANES and EXAFS analyses of dried solids, and sequential extraction can indicate the oxidation state and mineral association of solid-bound chromium. The entrained-liquor contribution is accessible only through the mass balance and soluble-ion tracers of Section 4.1.

4.3. Literature-Constrained Illustrative Consistency Checks

The diagnostic balances in Section 4.1 are most reliable when final Cr concentration in the mother liquor, product Cr content, cake moisture, wash ratio, wash-liquor composition and conservative tracers are reported together. Such complete datasets are rarely available. The following examples are therefore literature-constrained, order-of-magnitude consistency checks rather than source-apportionment validations. Even in this constrained form, the data identify whether entrainment is a significant contributor to product Cr, helping to determine which pathway should be prioritized for process improvement under the specific process conditions. The key is that the entrainment estimate is an upper limit (Section 4.1): where this limit falls well below the measured product Cr, entrainment can be ruled out as the dominant source even though the exact partitioning is uncertain.

Hu et al. (2021) [9] investigated NH₄VO₃ precipitation from a chromium-containing vanadate solution using NH₄HCO₃. In one experiment with an initial Cr(VI) concentration of 5000 mg L⁻¹, N/V molar ratio of 2.0, pH 8.0, 30 °C and 5 h, the V precipitation efficiency was 80.1%, and the calcined V₂O₅ product contained 0.61 wt% Cr. If this value is elemental Cr on a calcined V₂O₅ basis and Cr is retained during calcination, it corresponds to 6100 × 0.778 ≈ 4.7 × 10³ mg kg⁻¹ on a dry NH₄VO₃ precursor basis. The factor 0.778 is the theoretical mass yield for 2NH₄VO₃ → V₂O₅ + 2NH₃ + H₂O. This conversion is approximate because the reported Cr basis, precursor purity, hydration state, washing history and Cr retention during calcination were not fully constrained. Using Equation (4), $\mathsf{\theta } = 0.30$ and ${\mathsf{\rho }}_{\mathrm{ML}} = 1.20$ kg L⁻¹ give $V_{\mathrm{ML},0} = 0.357$ L kg⁻¹ dry product. Because the final Cr concentration in the mother liquor was not reported, the initial Cr(VI) concentration was used only as a rough proxy for the upper limit. With no washing and assuming the initial Cr(VI) concentration persists (a rough upper limit), the calculated entrainment-derived Cr would be 5000 × 0.357 = 1.8 × 10³ mg kg⁻¹, about 38% of the converted product Cr. If the final mother-liquor Cr did not increase substantially during precipitation, this upper-limit calculation suggests that entrainment might account for up to ~40% of the product Cr under the no-wash scenario. If this upper limit is representative, solid-associated pathways (adsorption, occlusion, co-precipitation, or crystallization-related transfer) would then contribute the balance; however, without knowledge of the final mother-liquor composition and washing conditions, this attribution remains tentative. For this system, process improvement should target these solid-phase mechanisms (e.g., controlled crystal growth) rather than intensifying cake washing alone.

Chen et al. (2023) [16] prepared ultra-pure NH₄VO₃ from a solution containing 0.05 g L⁻¹ Cr(VI) using controlled heterogeneous self-assembly crystallization. The final product contained <10 mg kg⁻¹ Cr. Using the same cake-retention parameters, the maximum entrainment-derived Cr would be ≈18 mg kg⁻¹. Because this upper limit already exceeds the measured product Cr, and only a moderate reduction in retained mother liquor (R_ML,W ≈ 0.56) is needed to reach <10 mg kg⁻¹, entrainment provides a sufficient and self-consistent explanation of the product Cr. The determination is therefore that entrainment can account for the observed contamination in full. Here, further reductions in product Cr would be most efficiently pursued through improved washing and dewatering. However, controlled crystal growth may also reduce Cr(VI) incorporation into the AMV product, as reported for the AMV system by Chen et al. [16]. Because the detailed AMV preparation conditions and the chromium speciation in the final product were not fully reported, this mechanistic interpretation is conditional on the assumption that the relevant chromium was present as Cr(VI).

Neither example described previously constitutes a validated source apportionment because Cr-specific key variables—final mother-liquor Cr, cake moisture, wash ratio, washwater composition and conservative tracer concentrations—were not reported. Nevertheless, both examples illustrate the intended diagnostic use of the mass balance: rather than merely confirming that entrainment is plausible, it can determine whether entrainment contributes significantly, which is what directs process improvement. In the first case, the low entrainment upper limit rules out entrainment as the dominant source and redirects effort toward solid-associated pathways; in the second, the limit on the product Cr identifies entrainment control as the effective lever. Future studies should report the variables listed above together with product Cr, as summarized in the reporting checklist in Section 6, so that this determination can be made quantitatively rather than relying on limit assumptions.

5. Mother-Liquor Treatment and Recycle Stability

5.1. Residual V/Cr Recovery and Cr(VI) Control

Post-precipitation mother liquors should be treated as active process streams rather than simple effluents. After AMV/APV precipitation, these may contain residual V, Cr, NH₄⁺, Na⁺, sulfate, chloride, carbonate, reductant by-products and other impurity metals. These components affect residual metal recovery, Cr(VI) discharge risk, salt accumulation, water balance and recycle stability.

Residual V in mother liquor is both a recovery opportunity and a recycle burden. The appropriate recovery method depends primarily on the residual V concentration and the NH₄⁺/Na⁺/SO₄²⁻ matrix. When residual V is relatively high (typically on the order of grams per liter), secondary precipitation by further addition of ammonium salt or acid hydrolysis is usually the most direct option [11,39], or the liquor may be returned to an upstream leaching or conditioning step [6,7]. At lower residual V (below roughly hundreds of mg L⁻¹), adsorption [40] or ion exchange [41] becomes more effective; solvent extraction [42,43,44] is preferred for large liquor volumes or where co-recovery of V and Cr is desired. In cyclic V/Cr processes, V recovery and Cr recovery are linked with liquor reuse and process circulation [11,39,40,41,42,43,44,45]. Therefore, residual V removal should be designed together with Cr control.

In V-first circuits, defined here as a process in which vanadium is precipitated or recovered before chromium treatment, Cr(VI) reduction followed by Cr(III) precipitation is commonly applied after AMV/APV removal [5,11,13,39]. A representative route uses sulfur-bearing reductants such as sodium metabisulfite, for which Cr(VI) reduction is rapid at pH 2–3 under mildly acidic conditions [4]; practical dosages are typically 1.3–1.5 times [46] the theoretical stoichiometric requirement to compensate for side consumption by dissolved oxygen and other oxidants. This stage should be controlled by Cr(VI) concentration rather than total Cr, because unreduced Cr(VI) cannot be removed by hydrolysis and will be carried into downstream recycle or discharge. Reduction should continue until residual Cr(VI) meets the applicable target (commonly ≤0.5 mg L⁻¹); a sharp drop in ORP accompanies the depletion of Cr(VI) and provides a convenient in-line indicator. Effective reporting should include reductant type, dosage, pH, ORP, reaction time, residual Cr(VI), total Cr and residual V [8,10,14].

Cr(III) precipitation must be controlled within a defined pH window. Cr(OH)₃ reaches its minimum solubility near pH 8–9; if pH is too low (below ~7), Cr(III) removal is incomplete; if pH is too high (above ~10), colloid formation, impurity co-precipitation or amphoteric redissolution as Cr(OH)₄⁻ may occur. Efficient removal of residual V before Cr precipitation can improve chromium recovery by reducing competitive co-precipitation and the formation of voluminous V–Cr gels that impair filtration [11,39,42].

Trace-impurity removal is useful when bulk V/Cr separation has been achieved but trace contaminants remain above the concentration limits required for liquor recycling, downstream salt recovery, or discharge. Methods include ion exchange, adsorption, or solvent extraction [40,41,43,44,47,48], which can lower residual V and Cr to the low mg L⁻¹ level or below. A relevant post-precipitation example is N263 co-extraction of V and Cr followed by selective stripping of chromium with NaOH and then of vanadium with HCl [42]. However, residual extractant or entrained organic phase must be assessed, since co-extracted sulfate or entrained organic phase can itself become a contaminant and disrupt downstream salt recovery.

For dissolved salts such as Na⁺, sulfate and chloride, selective V/Cr treatment alone is insufficient: these species accumulate during closed-loop recycle and must be managed by purge or salt recovery. Rather than treating the purge as an adjustable loss stream, the practical route is to recover the salt load directly. Comprehensive treatment of vanadium precipitation liquor can recover Cr, ammonia and salts [13], while fractional crystallization of Na₂SO₄ and (NH₄)₂SO₄ provides a direct route for treating salt-rich purge streams [49]. The fate of residual V and Cr(VI) during salt crystallization should be verified before integration, because both may report to the crystalline salt phase, remain in the mother liquor, or concentrate in a recycled side stream.

Beyond metal recovery, post-precipitation liquor management involves several secondary burdens: ammonia loss through volatilization, salt accumulation requiring periodic purge, Cr(III) hydroxide sludge formed from reductive treatment, and washwater consumption. Certain Cr-separation methods, such as lead–salt precipitation, also introduce secondary Pb contamination. Process design should balance metal recovery against these environmental and resource impacts.

5.2. Mother-Liquor Recycle and Accumulation Risks

Mother-liquor recycling can reduce water consumption and recover residual reagents, but recycled liquors differ chemically from fresh liquors. Residual V/Cr and various salts may accumulate over cycles, changing ionic strength, viscosity, vanadium activity, supersaturation, nucleation, filtration behavior and impurity transfer. A conceptual representation of mother-liquor recycling and associated accumulation risks is shown in Figure 2.

Figure 2. Schematic of the mother-liquor recycle loop and accumulation risks, showing the locations of makeup-water addition, purge withdrawal, salt recovery, trace-impurity removal, and the associated recovery and discharge pathways.

Because dissolved salts (such as Na⁺ and SO₄²⁻) and unreduced Cr(VI) accumulate in the absence of a proper bleed, salt accumulation can shift the solubility window, promote fine-particle nucleation, increase slurry viscosity and worsen filtration, ultimately raising Cr/Na/SO₄²⁻ incorporation into the AMV/APV product. Cr(VI) build-up also severely increases product contamination risk. The steady-state concentration of a non-reacting species is governed by the balance between its net input per cycle and its removal rate—whether by purge (bleeding a fraction of the liquor, which removes all dissolved species indiscriminately) or by selective salt removal (e.g., crystallization of accumulated Na₂SO₄/(NH4)₂SO₄ [49], which targets the salts while retaining the liquor). In industrial circuits, however, variable liquor volume, makeup-water addition, evaporation, washwater return and salt precipitation mean that no constant enrichment factor (steady-state concentration relative to the first-cycle feed) can be assumed; the accumulation level must therefore be determined from cycle-resolved measurements—tracking Na⁺, SO₄²⁻, Cl⁻, Cr(VI) and product-impurity trends over multiple cycles. An industrially credible recycle strategy must treat purge design as a central control variable, and should include residual V/Cr recovery, purge or crystallization for accumulated salts, and multi-cycle validation of product quality, as summarized in the reporting checklist in Section 6.

6. Integrated Diagnostic and Process-Control Framework

This framework provides an engineering sequence tailored to ammonium-rich circuits by linking speciation, entrainment, and mother-liquor recycling. It complements the equilibrium-based speciation and precipitation analyses discussed in Section 3.1, Section 3.2 and Section 3.3 with practical industrial indicators.

6.1. Reporting Checklist

The framework supports four practical decisions: whether product Cr is caused mainly by entrained liquor or solid-associated transfer; whether washing or crystallization control is the more relevant corrective action; whether Cr(VI) should be controlled before recycle; and how the purge stream should be managed for long-term operation, including possible salt recovery.

Credible chromium-control evaluation requires coordinated reporting of feed chemistry, Cr valence, precipitation, solid–liquid separation, product quality, mother liquor and recycle history (Table 2).

Table 2. Recommended reporting indicators for ammonium-rich vanadium precipitation circuits.

Task	Recommended Indicators
Feed diagnosis	Slag type, roasting/leaching route, pH, ORP, total V, total Cr, Cr(VI)/Cr(III)
Salt matrix	NH4+, Na+, SO42−, CO32−, Cl−, conductivity
Vanadium precipitation	pH trajectory, NH4+ dosage, temperature, time, seed, V precipitation efficiency
Chromium control	Cr(VI), Cr(III), total Cr, reductant, final pH, residual Cr(VI)
Solid–liquid separation	Particle size, filtration rate, wet/dry cake mass, cake moisture
Entrainment diagnosis	Cr concentration in the mother liquor, equivalent entrained liquor, wash ratio, wash efficiency, wash-liquor composition, V loss during washing, cake moisture, product Na and SO42−
Colloid diagnosis	Unfiltered/filtered Cr, 0.45 μm and finer filtration, turbidity, DLS or particle-size data (if available), zeta potential (if available), Fe/Al/Si, aging time, settling time or filtration resistance as practical measures
Product quality	Product basis, calcination basis, V grade, Cr or Cr2O3, Na, SO42−, Fe, Al, Si, phase, moisture
Mother liquor	Residual V, Cr(VI), total Cr, NH4+, Na+, sulfate, chloride, pH, ORP, conductivity, density
Recycle validation	Cycle number, recycle ratio, purge ratio, salt balance, product-impurity trend
Residue stability	Cr(VI) in solids, leaching toxicity, risk of Cr(III) reoxidation to Cr(VI)

6.2. Diagnostic Decision Workflow

The workflow in Figure 3 starts with stream identification and chemical diagnosis. Key streams should be characterized by pH, ORP, total V/Cr, Cr(VI)/Cr(III), NH₄⁺, Na⁺, sulfate, chloride, suspended solids and recycle history.

Figure 3. Diagnostic decision workflow for chromium control in ammonium-rich vanadium precipitation circuits. Dashed regions indicate four functional modules: feed and speciation diagnosis, precipitation and product-Cr diagnosis, mother-liquor treatment, and recycle and accumulation control. An intermediate solid–liquid separation (filtration and washing) step follows impurity/Cr(III) removal and precedes AMV/APV precipitation.

If V is sufficiently concentrated for efficient precipitation and Cr is mainly present as Cr(VI), a V-first AMV/APV precipitation route can be used, with emphasis on supersaturation control, crystal size, filtration, washing and entrainment. If product Cr is high, the entrainment balance, cake moisture, washing response and Na/S/Cl tracer data should be checked before assuming failure of chemical V/Cr separation.

If the mother liquor contains significant Cr(VI), residual V recovery, Cr(VI) reduction, Cr(III) precipitation or trace-impurity removal should be selected according to liquor composition and recycle route. Solvent extraction (e.g., N263 co-extraction of V and Cr followed by selective stripping [42]) is included as a recovery option specifically for large liquor volumes or where co-recovery of V and Cr is desired; for typical mother-liquor volumes and V-only recovery, secondary precipitation, adsorption, or ion exchange is generally more practical. “Significant” residual V here refers to concentrations at or above the grams-per-liter level, at which secondary precipitation or return to an upstream step is the more direct recovery route; below roughly a few hundred mg L⁻¹, recovery by adsorption or ion exchange is more effective and the liquor is otherwise recycled, as discussed in Section 5.1. During mother-liquor recycling, residual V/Cr, Cr(VI), NH₄⁺, Na⁺, sulfate, purge ratio and product-impurity trends should be monitored over multiple cycles. Salt accumulation or Cr(VI) build-up should trigger trace-impurity removal from the bleed stream, salt recovery or purge redesign. Where Cr(III) hydrolysis or Fe/Al/Si removal is applied before vanadium precipitation, the resulting solids must be separated and the cake washed before the liquor enters AMV/APV precipitation; carrying these hydroxide or gel solids into the precipitation stage reintroduces the colloid-capture and entrainment pathways identified in Section 4.2. Where these solids are colloidal rather than filterable, removal relies on controlling the precipitation pH window (near the Cr(OH)₃ solubility minimum at pH 8–9) and aging, as discussed in Section 5.1, rather than on conventional filtration alone.

6.3. Industrial Implications for Slag-Derived Vanadium Circuits

The preceding analysis has three direct implications for industrial practice, particularly for circuits processing high-chromium vanadium slags derived from the Panxi region in China. In this region, the Hongge vanadium–titanium magnetite is characterized by a distinctly high Cr₂O₃ content (0.4–1.1 wt%) [50], which is 5–10 times higher than that in surrounding deposits [3]. During subsequent converter smelting, most chromium in the feed is oxidized together with vanadium and enriched in vanadium slag [5,51]. Coarse vanadium slag from such high-chromium feedstocks may contain Cr₂O₃ in the range of 6–9 wt% or higher; for example, Li et al. [51] reported 6.32 wt% Cr₂O₃ in high-chromium vanadium slag, while Wen et al. [5] documented 9.45 wt% Cr₂O₃ in similar material and noted that this value is much larger than that in normal vanadium slag. This elevated chromium burden makes Cr control a particularly critical consideration in process design and operation.

First, upstream roasting determines Cr(VI) and salt load. The choice of sodium roasting, calcification roasting, ammonium-sulfate roasting or magnesiation roasting dictates not only vanadium extraction but also the initial Cr(VI)/Cr(III) ratio and the Na⁺/NH₄⁺/SO₄²⁻/Cl⁻ matrix entering the precipitation circuit [4,5,6,7,52,53,54,55,56]. A V-first strategy is most credible when the upstream process does not introduce an excessive soluble Cr(VI) burden that later accumulates in recycled mother liquor [8].

Second, ammonium vanadate precipitation should be optimized together with filtration and washing. Product chromium may correlate strongly with cake moisture, crystal size and washing efficiency, which are often optimized for throughput rather than chromium control. The entrainment balance and tracer approach in Section 4 provide a quantitative basis for deciding whether a product-Cr target requires lower cake moisture, improved washing, lower Cr concentration in the mother liquor or controlled crystal growth.

Third, mother-liquor recycling requires purge, Cr(VI) control and salt-balance management. Direct recycling can save reagents and water, but dissolved salts and unreduced Cr(VI) may accumulate. Where high water recovery or zero-discharge operation is required, fractional crystallization of Na₂SO₄/(NH₄)₂SO₄, Cr(VI) reduction–precipitation and/or extensive purification of bleed streams should be integrated with the main precipitation circuit.

6.4. Research Needs

Future work should move from single-pass precipitation evaluation toward source-resolved chromium accounting and recycle validation. Five priorities are suggested.

First, chromium speciation should be linked with operationally defined solid size fractions so that dissolved, colloidal and particulate chromium can be distinguished.

Second, product-Cr source apportionment should be validated using combined entrainment, washing and soluble-tracer tests, rather than inferred only from final product grade.

Third, AMV/APV crystallization, filtration and washing should be studied as coupled operations because crystal size, cake structure (porosity and permeability) and wash efficiency jointly determine impurity transfer.

Fourth, mother-liquor recycle should be evaluated over multiple cycles with explicit purge or salt-recovery strategies, so that residual V, Cr(VI) and accumulated salts are controlled together.

Finally, Cr-bearing residues generated during reduction or precipitation should be assessed for reoxidation and subsequent Cr(VI) leaching risk under realistic storage conditions. If independent characterization confirms that product chromium is predominantly associated with Cr-rich Cr(III) colloids, controlled oxidative redissolution could be investigated as a possible treatment option to transfer chromium into solution. Any resulting Cr(VI) would then require reduction and subsequent solid–liquid separation. The feasibility of this approach would depend on colloid composition, co-precipitated phases and compatibility with V recovery, and therefore requires targeted validation rather than being assumed as a general treatment.

7. Conclusions

Chromium control in ammonium-rich vanadium precipitation circuits is governed by speciation, crystallization, solid–liquid separation, and recycle effects. Cr(VI) may remain soluble during V-first AMV/APV precipitation, whereas Cr(III) may hydrolyze to form hydroxide precipitates or colloids that can enter products through colloid capture or association with impurity hydroxides. Therefore, chromium behavior cannot be evaluated reliably from total Cr removal or single-pass vanadium precipitation efficiency alone.

This review integrates Cr valence analysis, entrainment balances, washing diagnostics, soluble-ion tracers and simplified recycle inventory checks into a diagnostic framework for ammonium-rich precipitation circuits. Product chromium can be screened by comparing measured Cr with the amount expected from retained mother liquor and by using Na⁺, SO₄²⁻ and Cl⁻ as soluble-ion tracers. Because none of the chromium fractions is directly measurable and tracer-based estimates are subject to adsorption artifacts on high-surface-area precipitates, this screening yields semi-quantitative limits rather than exact source attribution; it should be corroborated by at least two tracers of contrasting adsorption affinity, together with washing-response behavior and independent solid-phase characterization. Where retained liquor does not fully account for product chromium, adsorption, occlusion or colloid capture provide testable alternatives to entrainment; for Cr(III)-bearing colloids in particular, control rests on pH-window selection and aging during precipitation, together with reductive precipitation or Fe/Al/Si coagulation for downstream Cr(VI) removal.

Post-precipitation mother liquor is an integral component of the recycle circuit and influences residual V recovery, Cr(VI) control, purge design, salt recovery and multi-cycle behavior. These aspects should be evaluated together with precipitation, filtration and washing. The framework is designed for process diagnosis and reporting consistency. Quantitative prediction requires cycle-resolved data and validation in real high-salt liquors.