The aqueous feed entering a tantalum-niobium solvent extraction circuit after hydrofluoric acid–sulfuric acid digestion of columbite-tantalite concentrate carries a burden of undigested gangue, precipitated fluoride salts, and colloidal hydrolysis products. Within this matrix, the ketone extractant—typically methyl isobutyl ketone (MIBK)—exhibits extraction selectivity for tantalum over niobium that depends on the speciation of TaF₆⁻ and NbOF₅²⁻ complexes, yet the realized selectivity in a continuous contactor is governed equally by hydrodynamic and interfacial phenomena. Suspended particulate matter above approximately 50 mg/L gravimetric loading reduces the effective interfacial area for mass transfer by competing with the organic phase for adsorption at the aqueous-organic boundary, a mechanism documented in mixer-settler extractor performance literature and confirmed by laser Doppler anemometry studies of droplet population dynamics. The selectivity factor β_Ta/Nb, defined as the ratio of distribution coefficients D_Ta/D_Nb, approaches equilibrium values of 100–400 in clean laboratory shake-out tests at 4 N HF concentration, but in production-scale contactors with turbid feed, the apparent β_Ta/Nb declines because the forward extraction rate of tantalum, being governed by interfacial complexation kinetics rather than bulk diffusion alone, is suppressed proportionally more than the extraction rate of niobium. Furthermore, the presence of fine particles below 5 μm in the feed leads to the formation of interfacial "crud," a heterogeneous accumulation of particulate matter, metal-loaded organic degradation products, and micro-emulsified aqueous droplets that accelerates solvent degradation by increasing residence time of the organic phase in contact with acidic aqueous solution and by providing catalytic surface area for MIBK hydrolysis to methyl isobutyl alcohol and acetone.

What Turbidity Threshold Compromises Interfacial Mass Transfer in Pulsed Columns?

Pulsed sieve-plate columns employed for tantalum solvent extraction operate with aqueous-continuous dispersion regimes where the Sauter mean droplet diameter of MIBK ranges from 0.8 to 2.5 mm under pulsing intensities of 20–35 mm stroke length at 60–120 pulses per minute. The dispersed phase holdup in these columns typically remains between 8% and 18% by volume, and the specific interfacial area available for mass transfer is inversely proportional to the Sauter mean diameter: a reduction in mean droplet size from 2.0 to 1.0 mm doubles the interfacial area per unit volume from approximately 1,500 m²/m³ to 3,000 m²/m³. When the clarified aqueous feed entering the bottom distributor of the column contains suspended solids above 25–30 NTU, measured according to ISO 7027-1:2016 using a 860 nm nephelometric source, measurable increases in coalescence inhibition occur because particles adsorb irreversibly at the organic-aqueous interface and form a rigid monolayer that prevents film drainage between approaching droplets. Published pilot-plant data for MIBK-HF systems indicate that phase disengagement time, measured as the time required for the dispersion band to collapse to half its initial thickness, increases from 60–90 seconds under clean-feed conditions to 300–600 seconds when the feed turbidity exceeds 50 NTU and the suspended solids concentration exceeds 100 mg/L as determined by ASTM D7060-12 membrane filtration. This fivefold to tenfold degradation in coalescence efficiency forces a reduction in pulsing intensity to maintain a stable dispersion band, which in turn increases the droplet diameter and reduces the interfacial area, thereby reducing the volumetric mass transfer coefficient k_L·a from its design value of 0.02–0.04 s⁻¹ to below 0.01 s⁻¹. The consequence for selectivity is asymmetric: the tantalum extraction rate, which is strongly influenced by the interfacial reaction of TaF₆⁻ with protonated ketone molecules, decreases sharply, while niobium co-extraction remains partially compensated by intraphase complexation with fluoride ions, leading to an apparent selectivity collapse during periods of elevated feed turbidity. The design of feed clarification systems preceding tantalum solvent extraction circuits involves a staged sequence of operations whose individual performance characteristics determine the achievable clarity in the clarified aqueous liquor. Gravity settling in thickener vessels equipped with rake mechanisms operating at 0.02–0.08 rpm provides primary solids removal, reducing suspended solids from digestion discharge concentrations of 10–100 g/L down to 500–2,000 mg/L in the thickener overflow. Secondary clarification using pressure leaf filters with polypropylene filter cloth of 25–50 μm nominal retention ratings further reduces suspended solids to 50–150 mg/L, achieving filtration rates of 0.5–2.0 m³/m²·h at differential pressures of 0.3–1.2 bar under continuous precoat renewal with diatomaceous earth or perlite body feed. Polishing filtration through cartridge filters rated at 1–5 μm absolute removes residual fines to below 10–20 mg/L, and in circuits processing recycled capacitor scrap or synthetic concentrates with elevated colloidal iron hydroxide content, final clarification through depth media filters or crossflow microfiltration membranes operating at 0.2–0.5 μm pore diameter is required to achieve turbidity below 10 NTU. The pressure drop across polishing filters increases non-linearly with solids loading, and typical replacement intervals for 1 μm absolute polypropylene cartridges range from 200 to 600 hours of operation depending on the feed solids burden and the effectiveness of upstream coagulation using high-molecular-weight anionic polyacrylamide flocculants dosed at 2–10 mg/L. Operational experience from continuous extraction campaigns documented in solvent extraction conference proceedings indicates that the total clarification train cost, including filter aid consumption, cartridge replacement, and maintenance labor, represents 3%–8% of the direct operating cost of the solvent extraction circuit, a figure that rises to 10%–15% when the ore feed contains elevated proportions of weathered minerals that generate colloidal fines upon acid digestion.

When Colloidal Silica Enters the Organic Phase Budget

The presence of colloidal silica in the clarified aqueous feed introduces a distinct failure mode in ketone-based tantalum extraction that operates through a mechanism separate from simple interfacial area reduction. During hydrofluoric acid digestion of silicate-bearing minerals, silicon dioxide dissolves partially as hexafluorosilicic acid (H₂SiF₆) and partially as a metastable colloidal suspension of hydrated silica particles with mean diameters between 10 and 200 nm. As the digested slurry is cooled from digestion temperatures of 85–95 °C to the extraction temperature of 30–40 °C, the solubility of hexafluorosilicic acid decreases, and polymerization of silicic acid monomers occurs, generating a fresh population of colloidal silica particles that pass through conventional clarification filters. When this colloidal silica-laden feed contacts the MIBK organic phase in the extraction mixer, the silica particles compete with tantalum fluoride species for interfacial adsorption sites. Published data for interfacial tension in MIBK-water systems demonstrate a reduction from 8–12 mN/m for the clean two-phase interface to below 3 mN/m when colloidal silica at concentrations exceeding 100 mg/L SiO₂ is present. This interfacial tension depression changes the droplet breakage and coalescence dynamics in the mixer, leading to the formation of a stable micro-emulsion with droplets below 50 μm that does not fully coalesce in the settler within the designed residence time of 15–30 minutes. The consequence is organic phase carryover into the aqueous raffinate and aqueous entrainment into the loaded organic, both of which degrade selectivity in subsequent stages: the aqueous carryover contaminates the loaded organic with niobium-rich aqueous droplets that report to the scrub section and strip section, while organic carryover into the raffinate represents a direct loss of tantalum-loaded extractant. Published data for this specific configuration is limited, but the general phenomenon of emulsion stabilization by colloidal silica in solvent extraction systems is documented in hydrometallurgical literature and in technical bulletins from mixer-settler equipment manufacturers. A monitoring program for feed clarity parameters in tantalum-niobium solvent extraction circuits relies on a combination of online instrumentation and laboratory reference methods to maintain the clarified feed within the operational envelope required for stable ketone selectivity. Online turbidity analyzers operating on the 90-degree scattered light principle with ISO 7027-1:2016 compliance provide continuous monitoring of feed turbidity at a frequency of one measurement per minute, with an accuracy of ±2% of reading or ±0.1 NTU, whichever is greater, across the range 0–200 NTU. These online measurements are calibrated against gravimetric total suspended solids determinations performed at four-hour intervals using ASTM D7060-12 membrane filtration through 0.45 μm mixed cellulose ester filters, dried at 105 °C to constant mass. The correlation between online turbidity and gravimetric suspended solids in HF-sulfuric acid leachate matrices is feed-specific: iron-rich feeds from weathered ore deposits show a mass-to-turbidity ratio of 0.35–0.62 mg/L per NTU due to the higher specific light-scattering coefficient of hydrous iron oxide flocs, while silica-dominated feeds from primary hard-rock operations exhibit a ratio of 0.15–0.25 mg/L per NTU. Particle size distribution measurements using laser diffraction according to ISO 13320:2020, in which the angular scattering pattern of a dispersed sample is recorded and inverted using Mie theory, are performed daily to track the population of fines below 5 μm, which are the most interfacially active and the most difficult to remove by conventional filtration. The d₉₀ diameter, defined as the particle size below which 90% of the particle volume resides, is maintained below 25 μm for operation of mixer-settler banks, while pulsed column circuits require a more stringent d₉₀ of 10–15 μm to prevent accumulation of interfacial crud at the plate perforations. The operational boundary established in several commercial facilities specifies that if the online turbidity exceeds 30 NTU for a cumulative duration of 60 minutes in any 8-hour shift, the extraction circuit is switched to internal recycle mode and the feed is diverted to a stand-by clarification vessel until the turbidity recovers to below 20 NTU. Published data for this specific configuration is limited; the thresholds cited reflect operating practices disclosed in solvent extraction conference proceedings and technical audits of base-metal and specialty-metal extraction facilities.

Filter Press Operations and Diatomaceous Earth Precoat Dynamics

The clarification of tantalum extraction feed using recessed-chamber filter presses equipped with polypropylene plates of 1,500 × 1,500 mm dimensions and hydraulic closing systems rated at 350 bar operating pressure follows a precoat and body-feed protocol that determines both the filtration rate and the achievable filtrate turbidity. The precoat layer, consisting of 1–2 kg/m² of filter-grade diatomaceous earth with a permeability of 1.5–3.0 darcy, is applied as a slurry in dilute hydrofluoric acid-compatible carrier at a flow rate of 1.0–1.5 m³/m²·h, forming a 3–5 mm filter cake that serves as the primary filtration medium. Body feed of additional diatomaceous earth is metered into the unfiltered feed stream at a ratio of 0.5–2.0 kg per cubic meter of feed to maintain cake permeability as fine particles accumulate and to extend the filtration cycle. Under these conditions, the press operates at an average filtration rate of 0.4–0.8 m³/m²·h for feeds containing 1,000–2,000 mg/L suspended solids, with cake discharge occurring when the chamber pressure differential reaches 7–8 bar or when the cake thickness compromises filtrate flow. The filtrate from a properly operated precoated filter press typically measures 5–15 NTU with gravimetric suspended solids below 20 mg/L, sufficient to protect polishing filters but not sufficient for direct feed to pulsed column contactors without additional polishing. The filter cake, which contains residual soluble tantalum and niobium fluoride species in the entrained liquid phase, is washed with dilute hydrofluoric acid (1–2 N HF) at a wash ratio of 1.5–2.0 bed volumes to recover the valuable metal values before disposal, and the wash liquor is recycled to the digestion circuit. The filtration equipment itself is constructed of carbon steel with a corrosion-resistant polymer lining of polyvinylidene fluoride (PVDF) or ethylene chlorotrifluoroethylene (ECTFE) at a minimum thickness of 2.5–3.0 mm, selected to withstand continuous exposure to hydrofluoric acid at concentrations up to 8 N and temperatures up to 50 °C, with the filter cloth being polypropylene needle felt of 850–1,000 g/m² basis weight. The chemical interaction between ketone extractants and particulate matter in the feed stream extends beyond physical interference to include catalytic acceleration of solvent degradation. MIBK undergoes acid-catalyzed hydrolysis in the presence of aqueous hydrofluoric acid, yielding methyl isobutyl alcohol and acetone as degradation products, with the hydrolysis rate constant at 30 °C in 4 N HF estimated from organic chemistry literature to be in the range of 10⁻⁷ to 10⁻⁶ s⁻¹ under homogeneous conditions. When the two-phase system contains suspended iron oxide or iron hydroxide particles at concentrations above 50 mg/L, the effective degradation rate increases because iron(III) species complexed at the particle surface act as Lewis acid catalysts for the keto-enol tautomerization that precedes nucleophilic attack by water. The accumulation of methyl isobutyl alcohol in the organic phase reduces the extraction selectivity for tantalum by two mechanisms: first, the alcohol acts as a competing solvating extractant for niobium oxyfluoride species, increasing the niobium distribution coefficient; second, the alcohol increases the mutual solubility of water in the organic phase, promoting aqueous droplet entrainment and altering the activity coefficient of protonated MIBK. Solvent quality control in commercial circuits therefore includes gas chromatographic analysis of the regenerated organic phase for methyl isobutyl alcohol content, with a specification limit of less than 0.5 wt% in circuits producing capacitor-grade tantalum oxide, and less than 1.5 wt% in circuits producing metallurgical-grade tantalum products. When the alcohol content exceeds these limits, the solvent is either redistilled at atmospheric pressure (MIBK boiling point 116 °C) or subjected to alkaline scrubbing with 5–10 wt% sodium carbonate solution, which hydrolyzes the alcohol back to the ketone via a reverse aldol pathway. Published data for this specific configuration is limited; the general reaction pathways are established in organic chemistry literature.

Crud Chemistry and Interfacial Tension Degradation

The interfacial accumulation of particulate matter, emulsified organic droplets, and precipitated metal salts—collectively termed crud—represents the most operationally disruptive consequence of inadequate feed clarity in tantalum-niobium solvent extraction. Crud formation initiates when suspended particles in the 1–10 μm size range become wetted by both phases and concentrate at the liquid-liquid interface in the settler. The surface chemistry of the particles determines their interfacial activity: silica particles with an isoelectric point near pH 2 acquire a negative charge in the highly acidic extraction medium (pH < 0) and interact preferentially with the protonated carbonyl group of MIBK, while iron hydroxide particles with an isoelectric point near pH 7–8 acquire a positive charge and interact with the fluoride anion within the TaF₆⁻ complex. The crud layer thus functions as a third pseudo-phase with its own extraction chemistry, retaining both tantalum and niobium in proportions that depend on the local acid concentration and the extent of particle surface coverage by adsorbed metal complexes. Continuous crud removal is required in production circuits, typically through vacuum skimming of the crud layer from the settler surface at intervals of 8–24 hours, with the skimmed crud treated by centrifugation to recover the entrained organic phase before the solids are returned to the digestion circuit or directed to waste. The recovered organic phase from crud centrifugation contains elevated concentrations of dissolved iron and silicon, and it must be washed with concentrated sulfuric acid (10–30 vol%) followed by water before being returned to the extraction circuit to prevent progressive deterioration of selectivity. Without this acid washing, the iron and silicon report back to the organic phase in the extraction stage and re-precipitate at the interface, accelerating crud re-formation. Published data from long-term monitoring of a commercial tantalum extraction circuit indicated that the crud accumulation rate, expressed as the volumetric increase in crud layer thickness per unit feed throughput, correlates with the square of the feed suspended solids concentration above a threshold of 20 mg/L, such that a doubling of suspended solids from 25 to 50 mg/L increases the crud accumulation rate by a factor of approximately four. The operational boundaries for feed clarity in tantalum-niobium solvent extraction are influenced by the specific contactor configuration and the quality specification of the final tantalum product. Mixer-settler banks consisting of multiple extraction stages (typically 8–12 for extraction, 6–8 for scrubbing, and 4–6 for stripping) tolerate higher suspended solids loadings than pulsed columns because the mechanical agitation in each mixer continuously re-disperses the organic phase and the settler provides a large interfacial area for crud accumulation without immediate disruption of flow patterns. Suspended solids concentrations up to 50–100 mg/L can be processed in mixer-settler circuits with acceptable, though reduced, performance, provided that the crud layer is regularly removed and the settled organic phase is sampled for metal concentration analysis at intervals not exceeding 2 hours. In contrast, pulsed columns, which offer countercurrent operation with reduced footprint and lower extractant inventory, demand feed clarity below 10–15 mg/L suspended solids and below 10 NTU turbidity because the narrow plate perforations (typically 3–5 mm diameter) and the vertical flow geometry make the column susceptible to blockage and emulsion pump-down when particulate matter accumulates at the plate-liquid interface. The use of annular centrifugal contactors, which achieve phase separation by centrifugal force at 500–2,000 g, permits operation with suspended solids up to 200 mg/L because the high centrifugal acceleration counteracts the interfacial adsorption force and promotes rapid disengagement, but the capital cost per unit throughput and the maintenance complexity of high-speed rotating seals restrict their application in tantalum circuits to specific augmentation scenarios such as the treatment of recycle streams or the recovery of tantalum from process bleed liquors. The selection among these contactor types for a given feed clarity condition involves a trade-off between capital expenditure, operating cost, and the acceptable reduction in selectivity or throughput.

Feed Clarity Parameter	Method	Operating Range	Selectivity Impact	Mitigation Response
Turbidity	ISO 7027-1:2016 (nephelometric, 860 nm)	5–10 NTU (optimal); >30 NTU (critical)	Interfacial area reduction up to 40% at 30 NTU	Divert feed to stand-by clarifier; increase precoat thickness
Suspended solids	ASTM D7060-12 (0.45 μm membrane, gravimetric)	10–20 mg/L (optimal); >100 mg/L (critical)	Crud accumulation rate follows squared dependence above 20 mg/L	Replace polishing cartridges; verify filter press discharge pressure
Particle size distribution	ISO 13320:2020 (laser diffraction)	d₉₀ < 25 μm (mixer-settler); d₉₀ < 10 μm (pulsed column)	Fines <5 μm dominate interfacial adsorption	Increase flocculant dosage; switch to depth media filtration
Colloidal silica	Molybdate-reactive spectrophotometry (APHA 4500-Si)	<50 mg/L SiO₂ (optimal); >150 mg/L (critical)	Interfacial tension depression from 8–12 to <3 mN/m	Increase digestion temperature; add aluminum sulfate coagulant at 20–50 mg/L
Emulsion stability	Phase disengagement time (in-house method)	60–90 s (clean); 300–600 s (contaminated)	Stable micro-emulsion reduces effective phase ratio	Increase settler residence time; apply crud skimming

The quality of the clarified aqueous feed entering the extraction circuit cannot be considered in isolation from the quality of the organic phase. Fresh MIBK as received from petrochemical suppliers typically contains 0.01–0.05 wt% water, 0.01–0.03 wt% methyl isobutyl alcohol, and trace quantities of acetone and other carbonyl by-products. Before introduction into the extraction circuit, the fresh MIBK is conditioned by contacting with the aqueous scrub liquor at a volumetric phase ratio of 1:1 to 1:2 for 30–60 minutes, which pre-equilibrates the organic with the acid composition of the circuit and pre-extracts some niobium. This conditioning step serves a dual purpose: it saturates the organic phase with water and acid so that volume changes during extraction are minimized, and it establishes the baseline niobium loading in the organic that the scrub stages must remove. When the conditioned organic phase is contacted with a clarified aqueous feed maintained at the specified clarity parameters, the selectivity of the extraction is reproducible within ±5% for batch-to-batch operation. When the feed clarity specifications are not met, the selectivity becomes erratic, with standard deviations of ±15% to ±30% across successive operational shifts. This variance propagates through the downstream stages, causing fluctuations in the niobium content of the final tantalum oxide product that may push the material outside the specification limit of less than 0.01 wt% Nb for capacitor-grade tantalum oxide, requiring additional purification passes and reducing overall circuit capacity. The economic impact of such clarity-induced variability is substantial: a 5% reduction in effective selectivity translates to a disproportionate increase in scrubbing reagent consumption because the scrubbing response curve is non-linear with respect to niobium loading.

Standard / Code	Title / Scope	Application to Feed Clarity	Testing Frequency
ISO 7027-1:2016	Water quality — Determination of turbidity — Part 1: Quantitative methods	Online and grab-sample turbidity of clarified feed; 90° scattered light, 860 nm source	Continuous (online); 4-hourly (grab)
ASTM D7060-12	Standard Test Method for Determination of Suspended Solids in Water by Membrane Filtration	Gravimetric suspended solids of feed and thickener overflow; 0.45 μm MCE filter, 105 °C drying	4-hourly during active extraction; daily during recycle
ISO 13320:2020	Particle size analysis — Laser diffraction methods	d₁₀, d₅₀, d₉₀ of suspended particles in clarified feed; Mie theory inversion	Daily
ISO 13319:2007	Determination of particle size distributions — Electrical sensing zone method	Verification of fines <5 μm when laser diffraction indicates bimodal distribution	Weekly or as required
APHA 4500-Si D	Standard Methods — Silica, molybdate-reactive method	Colloidal and reactive silica in clarified feed; spectrophotometric at 410 nm	Daily when silica-bearing feeds processed
ISO 17894	Solvent extraction — Vocabulary and general principles	Reference for extraction terminology and phase disengagement characterization	As required

The clarification equipment train for a tantalum-niobium solvent extraction circuit processing 500–2,000 tonnes per year of contained Ta₂O₅ operates under conditions that test the corrosion resistance of every wetted component. Hydrofluoric acid at concentrations of 4–8 N attacks most metals, glass, and ceramic materials, restricting the choice of construction materials to polymers such as PVDF, ECTFE, polytetrafluoroethylene (PTFE), and high-density polyethylene (HDPE) for wetted surfaces. The thickener vessel for primary solids removal is typically constructed of rubber-lined carbon steel with a minimum rubber lining thickness of 5 mm over the entire wetted surface, and the rake mechanism is fabricated from Hastelloy C-276 or coated with PTFE to withstand both mechanical abrasion and chemical attack. The pressure leaf filter is manufactured with a PVDF-lined carbon steel shell and a filter leaf assembly of polypropylene mesh over a polypropylene framework, with the precoat and body-feed slurry prepared in a separate agitated vessel equipped with a HDPE lining and a mechanical seal pump rated for hydrofluoric acid service. The polishing filter housings, constructed of PVDF or ECTFE with a design pressure of 10 bar at 50 °C, contain polypropylene depth filter cartridges with a micron rating of 1–5 μm absolute, and differential pressure indicators across the housing provide the signal for cartridge replacement. The clarified feed storage tank, with a capacity equal to 8–16 hours of circuit feed rate, is fabricated from HDPE or cross-linked polyethylene with ultraviolet stabilization, and it is equipped with an agitator to maintain homogeneity without introducing additional shear that could break apart flocculated solids. The entire clarification train is designed for service with feed streams containing up to 8 N HF, 6–10 N H₂SO₄, 10–100 g/L dissolved iron, 1–20 g/L dissolved aluminum, and variable concentrations of titanium, tin, tungsten, and rare-earth metals that depend on the ore mineralogy. The relationship between feed clarity and ketone selectivity is further modulated by the aqueous phase chemistry of fluoride complexation. In fluoride-bearing solutions, tantalum and niobium exist as a distribution of fluoro- and oxyfluoro-complexes whose relative proportions depend on the fluoride-to-metal ratio, the hydrogen ion concentration, and the presence of sulfate ions from the sulfuric acid co-solvent. Tantalum is extracted by MIBK as the hexafluorotantalate anion TaF₆⁻ associated with a protonated ketone molecule, while niobium is extracted predominantly as the oxypentafluoroniobate anion NbOF₅²⁻ at low HF concentrations and as the hexafluoroniobate NbF₆⁻ at high HF concentrations. The selectivity for tantalum over niobium is highest at intermediate HF concentrations of 4–6 N, where niobium exists primarily as the less extractable oxyfluoride species, and it declines at higher HF concentrations where niobium is fully converted to the hexafluoro species and competes more effectively for the protonated extractant. Suspended particulate matter in the feed influences this equilibrium speciation by providing adsorption surfaces that preferentially bind one species over another. Iron hydroxide particles, for example, exhibit a strong affinity for the anionic fluoride complexes through electrostatic interaction with the positively charged iron surface in acidic solution, and this adsorption removes a fraction of the total dissolved tantalum from the aqueous phase available for extraction. The adsorbed tantalum is not irreversibly lost but is temporarily sequestered at the particle surface, where it remains available for extraction only at a reduced effective concentration. This reduction in free tantalum concentration shifts the extraction equilibrium unfavorably, decreasing the apparent distribution coefficient for tantalum and thereby decreasing the observed selectivity factor. Published data quantifying the extent of tantalum adsorption onto particulate matter in HF-SO₄ media is limited, but the general phenomenon of metal adsorption onto suspended solids in hydrometallurgical leachates is well documented. Operating experience from continuous tantalum extraction campaigns demonstrates that the impact of feed clarity on process performance is neither linear nor predictable from simple measurements of turbidity alone. Batch-to-batch variance in feed quality, even within the same ore deposit, arises from the heterogeneous distribution of minerals within the deposit, from differences in blast fragment size distribution that affect digestion efficiency, and from fluctuations in the digestion reagent addition rates. The suspended solids in the digested slurry exhibit a particle size distribution that shifts toward finer sizes as the proportion of weathered mineral matter increases, and this shift in particle size distribution changes the interfacial behavior of the solids even when the total suspended solids concentration remains constant. Two feed batches with identical turbidity readings of 20 NTU and identical gravimetric suspended solids of 30 mg/L can nevertheless exhibit different selectivity impacts if one batch contains predominantly angular quartz particles larger than 10 μm and the other contains predominantly sub-micron iron hydroxide colloids. The quartz particles, being relatively large and chemically inert, settle rapidly and have limited interfacial activity, while the iron hydroxide colloids, being positively charged and highly surface-active, concentrate at the liquid-liquid interface and significantly impair coalescence. This distinction necessitates the use of particle size distribution measurement and zeta potential determination as diagnostic tools for troubleshooting clarity-related selectivity problems in addition to the routine monitoring of turbidity and suspended solids. The zeta potential of suspended particles in the highly acidic extraction feed is typically measured by electrophoretic light scattering according to ISO 13099-1:2012, and it provides a direct measure of the electrostatic driving force for interfacial adsorption: particles with zeta potentials exceeding ±20 mV exhibit strong interfacial activity, while particles with zeta potentials below ±5 mV are largely inert. Published data for this specific configuration is limited, but the general relationship between particle zeta potential and interfacial adsorption is well established in colloid science literature. The operational response to a detected feed clarity excursion in a tantalum extraction circuit follows a defined sequence of corrective actions that progresses from minimal intervention to full circuit shutdown depending on the severity and persistence of the excursion. A transient turbidity increase to 30–50 NTU lasting less than 30 minutes, typically caused by a short-term disturbance in the upstream clarification train, is managed by increasing the coagulant dose to the clarifier, increasing the body feed rate to the filter press, and reducing the extraction circuit throughput by 10–20% to restore phase disengagement stability. A sustained turbidity increase above 50 NTU lasting more than 60 minutes triggers a more aggressive response: the extraction circuit is switched to internal recycle, the feed is diverted to a stand-by clarified feed tank, and the polishing filters are replaced or backwashed. If the turbidity remains above 100 NTU for more than 4 hours, the root cause is typically traced to a failure in the primary clarification equipment—a broken filter leaf, a spent precoat, or an overloaded thickener—and the circuit remains on internal recycle until the clarification train is restored to its design performance. During internal recycle operation, the organic phase in the contactor is maintained in circulation through the scrub and strip sections to prevent stagnation and to preserve the solvent quality, while the aqueous feed is held in the clarified feed storage tank with agitation to prevent settling and re-suspension of solids. The selectivity of the extraction process during recovery from a clarity excursion must be re-established by adjusting the organic-to-aqueous flow ratio and the scrub acid concentration, as the presence of residual crud in the settler and residual solids in the pumped feed requires a period of 2–6 hours of operation under reduced throughput before the circuit reaches its design selectivity specifications. Acid consumption in the digestion and extraction circuit is directly influenced by the clarity of the feed. When suspended solids are present in the clarified feed at concentrations above 50 mg/L, the effective acid concentration at the extraction interface is reduced because protons are consumed by surface reactions with the solid phase: hydroxyl groups on the surface of silica and iron oxide particles are protonated, and carbonate minerals such as calcite and dolomite dissolve in the acid medium, consuming protons and releasing carbon dioxide. This localized acid depletion at the particle surface creates a microenvironment with reduced HF activity, which shifts the tantalum speciation toward less extractable hydroxy-fluoride complexes and reduces the extraction rate. The influence of this acid depletion is most pronounced in the boundary layer immediately adjacent to the particle surface, where the acid concentration can be reduced by 0.5–1.0 pH units relative to the bulk solution, and it extends into the surrounding aqueous phase to a distance governed by the thickness of the Nernst diffusion layer, typically 10–50 μm in agitated contactors. Because the tantalum extraction rate depends on the available HF concentration, this surface acid depletion creates zones of reduced extraction efficiency around each suspended particle, and the cumulative effect scales with the total particle surface area rather than the total particle mass. This explains the observation that fine colloidal particles, which contribute little to the total suspended solids concentration but provide a large specific surface area, can exert a disproportionately large negative impact on extraction kinetics. The measurement and control of feed clarity in tantalum-niobium solvent extraction represents a critical process variable whose influence extends from the extraction stage through the scrubbing, stripping, and solvent regeneration stages. In the scrub section, where the loaded organic phase is contacted with 1–2 N HF and 1 N H₂SO₄ to selectively back-extract co-extracted niobium while retaining tantalum in the organic phase, the presence of entrained aqueous droplets from the extraction settler, which are enriched in suspended solids and dissolved niobium, reduces the scrubbing efficiency and increases the number of equilibrium stages required to achieve the target niobium rejection. In the strip section, where tantalum is recovered from the organic phase by contacting with water or dilute acid at a volumetric phase ratio of 1:1 to 2:1, the presence of suspended solids in the strip solution generates interfacial crud that interferes with the complete recovery of tantalum and contributes to organic phase losses. The solvent regeneration section, which comprises a distillation column or an alkaline wash system, is also affected because solids carried over from the extraction circuit foul the reboiler surfaces and reduce the heat transfer efficiency, increasing the energy consumption per unit of regenerated solvent. The cumulative effect of feed clarity across all these unit operations justifies the capital investment in comprehensive clarification equipment and the operational attention to maintaining the clarified feed within the specified clarity envelope. Published data for this specific configuration is limited; the general design principles are drawn from solvent extraction plant operations described in standard references on hydrometallurgical process engineering. The use of alternative ketone extractants for tantalum-niobium separation introduces additional clarity-related considerations that differ from those applicable to MIBK. Cyclohexanone, methyl ethyl ketone (MEK), and diisobutyl ketone (DIBK) have been evaluated in laboratory and pilot studies for tantalum extraction, and each exhibits a characteristic sensitivity to feed clarity that is related to its molecular structure and its interfacial behavior. Cyclohexanone, with a density of 0.947 g/mL and an aqueous solubility of approximately 8 g/L at 25 °C, is more polar than MIBK and extracts water more readily, making it more susceptible to emulsion formation when the feed contains surface-active particulates. DIBK, with a density of 0.805 g/mL and a low aqueous solubility of less than 0.5 g/L, is less polar and less susceptible to emulsion formation but exhibits slower phase disengagement in the presence of fine particles because its higher viscosity limits the drainage rate of the aqueous film between approaching droplets. The commercial preference for MIBK in tantalum extraction is based primarily on its favorable compromise between selectivity, phase disengagement rate, and cost, but the clarity tolerance of MIBK is not unlimited, and circuits using MIBK require the same level of feed clarification as circuits using alternative ketones. Published data for the comparative clarity tolerance of alternative ketone extractants in tantalum service is limited; the general trends are inferred from solvent extraction fundamental studies and from analogous applications in the pharmaceutical and fine-chemical industries. The interface between the clarified aqueous feed and the MIBK organic phase in the extraction mixer operates under conditions of high shear and rapid mass transfer, and the behavior of suspended solids at this interface determines the extent to which ketone selectivity is degraded. In a typical mixer-settler extraction stage for tantalum, the mixer residence time is 2–5 minutes, the impeller tip speed is 2–4 m/s, and the phase ratio (organic/aqueous) is maintained at 1:1 to 2:1 depending on the metal loading of the feed and the distribution coefficient. Under these conditions, the energy dissipation rate in the mixer ranges from 1 to 5 W/kg, generating a droplet size distribution with a Sauter mean diameter of 100–500 μm in the mixer and a coalescence rate in the settler that depends on the interfacial tension, the continuous phase viscosity, and the presence or absence of interfacial contaminants. Suspended solids in the feed are subjected to intense shear in the mixer, which can fragment particle agglomerates and release previously encapsulated metal species, but the same shear also exposes fresh particle surfaces to the organic phase, promoting adsorption and the formation of particle-stabilized emulsions. The balance between these competing effects depends on the particle size distribution and the surface chemistry of the solids, and it varies from one feed source to another. In feeds derived from primary hard-rock ores, the solids consist predominantly of quartz, feldspar, and refractory tantalum minerals that remain chemically inert in the extraction environment, while in feeds derived from recycled capacitor scrap or secondary sources, the solids may include metallic tantalum fines, tantalum oxide dust, and combustion residues that are chemically reactive and more prone to interfacial accumulation. The operational response to each feed type differs accordingly, and the clarification train must be adapted to the specific particle populations present.