Suspended Anode Slimes in Copper Electrorefining: What Keeps Them in the Electrolyte, and What That Means for Filtration

Short answer: the slime your electrolyte filter has to deal with is not the slime that lands in the cell bottom. One published cell study splits the two populations by size, reporting suspended slimes as mostly below 9 to 10 microns and settled slimes as commonly above it. The boundary between them moves with cell temperature and with the arsenic, antimony and bismuth balance in the electrolyte, which means part of your filtration duty is manufactured upstream of the filter and is partly within your control.

Anode slime is a recovery stream. It carries the gold, silver, selenium and tellurium that came in with the anode, and the questions that follow from that are how much of it there is and how to get the precious metals out. Those are real questions. They are not the filtration question.

The filtration question is narrower: of everything liberated from the anode, how much never reaches the cell bottom at all, and what decides that fraction? That is the material your clarification filter sees, and the sources below indicate it behaves differently from the material in the slime pit. For the duty in general, and where clarification sits in a tankhouse, start with our guide to electrolyte filtration for copper refineries.

The two slime populations, and why only one of them is your duty

Work by Zeng, Free and Wang, published in the Journal of The Electrochemical Society in 2016, sorted anode slime by where it ended up rather than by what it contained. It reports that the sizes of suspended slimes are mostly smaller than 9 to 10 microns, and the sizes of settled slimes are commonly larger than 9 to 10 microns.

The study reports the two populations as largely separated by size, and the hedges are its own: "mostly" below and "commonly" above, which is an admission that the distributions do overlap by an amount it does not quantify. The slime that is harvested and shipped to precious-metals recovery is largely a coarse population, and the slime circulating through your filter is largely a fine one.

A companion paper from the same group modelled the transport side. It represented suspended slime as 2 micron particles at a modelled particle density of 8000 kilograms per cubic metre, and set out why those particles stay where they are anyway: suspended slimes have small diameters and low settling velocities, allowing them to remain suspended in electrolyte by upward electrolyte flow. The same circulation that the cell needs, to keep copper ion concentration up at the cathode film and sulfate at the anode, is what holds the fines in the circuit.

That paper models a looping flow pattern within every anode-cathode gap, with upward flow along the cathodes, a path that would carry a suspended particle to a growing deposit. The first paper adds a measured association: the weight of suspended slimes is positively correlated with the bismuth level in the cathode.

So the fine fraction is not a housekeeping nuisance. It is the fraction that reports to your product.

For what those particles actually are, and how they behave against a filter medium, see our note on specifying filter cloth for a copper electrolyte duty, which covers the particle characterisation and the septum selection that follows from it.

What the settling boundary depends on, and the result that runs the wrong way

If the split between suspended and settled is a settling-velocity question, then anything that changes particle size changes the split. Temperature does, through coalescence at the anode surface.

That study reports peak adhesion temperatures of 55 °C, 60 °C, 65 °C and 70 °C depending on the anode type, and reports that above the peak the coalesced large particle clusters become unstable and fall from the anode surface to settle at the bottom. Material that has grown into clusters is material that is no longer circulating.

The cathode impurity data in the same study runs in the opposite direction to the intuition that a cooler cell is a gentler one. For high-impurity anodes, bismuth in the cathode measured 1.890 ppm at 25 °C and fell below 0.100 ppm at 60 °C and above. For low-impurity anodes it fell from 0.620 ppm to 0.100 ppm at 85 °C. Running colder did not produce a cleaner cathode in that work. It produced a dirtier one. That is consistent with fine particles coalescing and dropping at the higher temperatures rather than circulating, though we are citing the study for the measurement and not for that explanation.

Two cautions before anyone reaches for the temperature setpoint. This is a laboratory and pilot-scale result on specific anode types, not a plant trial, and the peak adhesion temperature is anode-dependent, which is the study's own point. And cell temperature in an operating tankhouse is set by current efficiency, additive behaviour, energy cost and anode passivation long before anyone considers its effect on slime settling. The useful reading is not "run hotter". It is that the suspended solids load looks to be at least partly a function of how the cell is run, not of anode grade and throughput alone.

That is worth knowing when a filtration duty is derived from a nameplate cathode tonnage.

The electrolyte compositions used in published cell studies sit in a narrow band, and it is worth being precise about what they are, because none of them is a plant assay. A Journal of the SAIMM study made up a solution from sulfate salts chosen to simulate one refinery's practice: copper at 45 g/L, sulfuric acid at 220 g/L and nickel at 15 g/L, at 70 °C. A 2021 study worked on a waste sulfate solution drawn from a commercial copper electrolysis, at copper 32.5 g/L and nickel 20.5 g/L, with nickel deliberately raised and copper lowered against the commercial values. The modelled pilot cell above ran at copper 45 g/L and sulfuric acid 180 g/L at about 62 °C.

The temperatures are the part with any claim to commercial practice, and even there the attribution is uneven. The 2021 study states 63 plus or minus 2 °C and 250 A/m² as the commercial operating condition, and one trade source gives 60 °C. The 70 °C above is a laboratory setpoint, not a stated plant figure. Across the sources the band is roughly 60 to 70 °C, and that band sits on top of the peak adhesion temperatures the 2016 study identified, which is part of why the effect is not academic.

The fine fraction you cannot settle out

Coalescence explains the slime that is mechanically liberated from a solid. It does not explain the part of the fine population that the sources describe as precipitating rather than eroding.

Arsenic, antimony and bismuth in the electrolyte are linked in the published work to the formation of what that literature calls floating slimes, and to the precipitation of antimony and bismuth arsenates. The sources we could reach do not locate where in the cell that precipitate forms, so neither will we. What they do state plainly is the consequence: inadequate control results in floating slime and insoluble arsenates, contributing to high nodulation of the copper cathode and impacting its quality and overall production efficiency.

The variable usually pointed at is the anode, and that is not wrong: one study notes that the so-called mineralogy of the anode directly affects the passivity of the anode, the formation of suspended slime and the deposition of slime. Where a refinery does not control its own smelter, though, anode chemistry is closer to an inherited input than an operating lever, and the electrolyte balance is the part the tankhouse itself operates.

This part of the load is described in the sources as precipitating rather than eroding off a solid, which is not something the adhesion and coalescence mechanism above addresses. The published lever for it is electrolyte chemistry control, which is the ratio discussed next.

The control ratio is published two ways, and we could not reconcile them

Here the literature stops being helpful in the way you would want, and it is worth being direct about it rather than picking whichever number suits.

The control target both of our sources reach for is the ratio of arsenic to the sum of antimony and bismuth. Two current peer-reviewed papers state it, both accessed for this article, and they do not agree.

• A 2023 paper in the Journal of Cleaner Production states that the molar ratio As/(Sb+Bi) must be maintained within the range 20 to 35, to prevent the occurrence of floating slimes at the lower value and precipitation of arsenates at the higher value.

• A 2025 paper in Sustainable Chemistry states that the industry enforces a molar fraction relation, expressed as MFR = As/(Sb+Bi), ranging between 2.9 and 3.3.

Same expression, same purpose, ranges an order of magnitude apart. Both papers cite prior literature for the figure rather than measuring it, and neither defines the term in enough detail to convert one into the other. We could not reconcile them from the published sources, and we are not going to assert that one of them is wrong.

We considered working the ratio out from the only composition above that reported both arsenic and antimony, and discarded it: that composition is a waste stream from a purification circuit rather than circulating tankhouse electrolyte, its bismuth was not reported, and a single non-comparable number would not have told anyone which paper to believe. It would have looked like evidence without being any.

The practical point stands without it. A ratio you cannot pin down from the literature is not a design input for a filter. If you are sizing a clarification duty, the number that matters is what your own electrolyte assay says, run on your own circuit, at your own anode grade. The published ratios are useful for understanding the mechanism and for framing a conversation with your own metallurgists. They are not a substitute for the assay.

The cathode is part of the filtration circuit

One more coupling deserves a mention, because additive dosing and solids removal sit in different parts of most specifications.

Additive control and solids control are not independent. The SAIMM paper on electrolyte additives, citing earlier work by Safizadeh and co-workers, notes that when additives are inadequately controlled the cathodes contain excessive nodules, dendrites and anode slimes entrapped on the surface, and that the resulting rough surfaces cause anode slimes and suspended solids to adhere and entrap onto the cathode surface, reducing the conductivity and ductility of the deposit. The nodular and dendritic growths also cause short circuits and decrease current efficiency.

Read together, those two sentences describe a rough deposit as a better collector of suspended solids than a smooth one. Our reading of that, and it is a reading rather than a published finding, is that the same solids concentration should cost you more in a cell whose additive dosing has drifted, and therefore that a tolerable suspended solids figure is unlikely to be a fixed plant constant. It would depend on the surface you are growing.

For context, the gap this all sits in is large: a fire-refined anode at 99.6 percent copper is refined to 99.999 percent, with impurities below 20 mg/kg. There is very little room in that specification for entrained particles.

What this changes about specifying the filter

Five things follow, and none of them require a number we do not have.

Specify against an assay, not a tonnage. A clarification duty derived from cathode output assumes a fixed relationship between production and suspended solids. The relationship is not fixed. Ask for the electrolyte assay and the circulation rate.

Ask which population the sample represents. A sample drawn from a settling tank, a slime pit or a launder after a quiescent section is not the same material as a sample from the circulating stream feeding the filter. Given how the two populations separate by size, the sampling point changes the answer.

Expect a fine duty with a low settling velocity despite a high particle density. A feed that is largely below 9 to 10 microns, with a component the sources describe as precipitating rather than eroding off a solid, is not a duty for a coarse septum. As we set out in the filter cloth piece, in a precoat duty the retention is done by the filter aid and the cloth's job is to hold the precoat and release the cake, which is why precoat is the mode that fits this feed. We have written separately on precoat and body feed and how the two divide the work.

Do not assume the load is steady. Anode changes, campaign changes, a smelter feed change and a drift in additive dosing all move it. Size for the range, not for the average.

Ask what the specification is actually protecting. It is cathode chemistry, and the evidence chain for that is specific: suspended slime weight correlates with cathode bismuth, and floating slimes contribute to nodulation. The clarification filter is a cathode-quality control point.

For the equivalent exercise on the solvent extraction side, where the solids arrive from the orebody rather than being created in the cell, see our piece on suspended solids in Copperbelt PLS. The contrast is the useful part: in an SX-EW circuit the feed is handed to you, and in an electrorefinery a meaningful part of it is made by the way you run the cells.

What we could not establish

Being straight about the gaps is part of the specification.

We found no published figure for suspended solids concentration in circulating electrorefining electrolyte, in mg/L, from any source we could reach. The industry survey work that would most likely carry it, the periodic world tankhouse operating data surveys, sits behind paywalls. We are not going to estimate one. If you have that number for your own circuit, you already have the most useful input to a filter specification, and it is worth more than anything published.

The 9 to 10 micron boundary comes from one research group. It is a specific, well-documented result and we have attributed it as such rather than presenting it as an industry constant.

Where Sharpenn fits

We build pressure filters for electrolyte clarification in non-ferrous electrolytic refining, in precoat and body-feed modes, using a flux-calcined diatomite filter aid.

~1,000 units installed worldwide across 460+ installations (all applications); 25 units in copper refining; capacities 5 to 350 m³/hr; repeat orders after 20 to 25 years.

Those 25 copper units run across India, Africa (DRC, Zambia) and Europe (Germany). The repeat orders are the part we would point a specifying engineer at: refineries came back to us after 20, 22 and 25 years, with the original filters still in service.

On the points above:

• The septum is a disposable filter cloth, a planned consumable that the customer replaces and that we can supply.

• Separately, we can supply filter bags for any make of filter, so an existing installation does not have to be one of ours for us to be useful to you.

• Precoat dosing runs at roughly 1 kilogram of filter aid per square metre of filtration area as a general rule of thumb. Right-sizing the precoat and tuning the body-feed rate to the actual slime load is where filter-aid cost is reduced, which is precisely why the load characterisation above matters commercially and not only technically.

• A cake discharge cycle typically costs zero to one hour of downtime, and we can provide hydraulic open and close operation on the horizontal configuration.

• We have designed to AS 1210 and designed and built to ASME VIII Division 1. Where a destination country requires certification, it is arranged through a reputed certification body per project and the equipment supplied certified.

If you are sizing or re-sizing a clarification duty, send us your electrolyte assay and your circulation rate. We will tell you what we would put in front of it, and what we would want measured first.

Frequently asked questions

What is the difference between suspended slimes and settled anode slimes?

They are largely different size populations, not the same material in two places. A study published in the Journal of The Electrochemical Society in 2016 found that suspended slimes are mostly smaller than 9 to 10 microns while settled slimes are commonly larger than 9 to 10 microns. The coarse population reports to the cell bottom and is harvested for precious-metals recovery. The fine population stays in circulation, and it is the one your electrolyte clarification filter has to remove.

Why do suspended slimes stay in the electrolyte instead of settling?

Because they are small enough that their settling velocity is low compared with the upward flow in the cell. The same 2016 work modelled suspended slime as 2 micron particles and described them as having small diameters and low settling velocities, allowing them to remain suspended by upward electrolyte flow. The electrolyte circulation that the cell needs for mass transfer at the electrode faces is also what keeps the fines aloft, which is why circulation rate and clarification duty have to be discussed together.

Do suspended slimes actually affect cathode quality?

Yes, and there is measured evidence for the association: the same study reported that the weight of suspended slimes is positively correlated with the bismuth level in the cathode. A companion paper modelled a transport path that would explain it, a looping flow within each anode-cathode gap with upward flow along the cathodes, carrying particles onto the growing deposit. Separately, published work on electrolyte purification notes that inadequate control of arsenic, antimony and bismuth results in floating slime and insoluble arsenates, contributing to high nodulation of the copper cathode.

Does running the tankhouse hotter reduce cathode contamination?

The published cell data points that way, though it is not a recommendation to change a setpoint. For high-impurity anodes, cathode bismuth measured 1.890 ppm at 25 °C and fell below 0.100 ppm at 60 °C and above. That is consistent with particles coalescing and settling instead of circulating, though the study is cited here for the measurement, not for that explanation. That is laboratory and pilot-scale work on specific anode types, and the peak adhesion temperature is anode-dependent. In an operating tankhouse, temperature is set by current efficiency, additive behaviour, energy cost and anode passivation, so treat this as an explanation of why your solids load varies, not as an operating instruction.

What As/(Sb+Bi) ratio should a copper refinery hold?

We cannot give you one from the literature with confidence, and you should be wary of anyone who does. Two current peer-reviewed papers state the control ratio for the same purpose and disagree by an order of magnitude: one gives a molar ratio of 20 to 35, the other gives 2.9 to 3.3. Both cite prior work rather than measuring it, and neither defines the term closely enough to convert between them. The practical answer is that this ratio is a plant control target to be set with your own metallurgists against your own assay, and it is not a design input for sizing a filter.

How do I specify an electrolyte clarification filter for an electrorefinery?

Start from the electrolyte assay and the circulation rate, not from cathode tonnage, because the relationship between production and suspended solids is not fixed. Check where the sample was drawn, since a sample from a settling tank or slime pit represents the coarse population rather than the fine one the filter sees. Expect a fine duty with a component the sources describe as precipitating rather than eroding off a solid. That is the case for running it as a precoat duty, where the filter aid does the retention and the cloth holds the precoat. Size for the range across anode campaigns and additive conditions rather than for the average.