Fluorite: staged cleaning under close temperature and reagent control

Say a mine owner asks, 'Can you build me a fluorite flotation plant?' The honest first question is not about tank cells or reagent suites. It's about the offtake contract. What grade does it call for? Acid, ceramic or metallurgical? That answer changes the entire circuit. Fluorite flotation is not one process, it's a family of flowsheets divided by the purity you have to sell.

Fluorite flotation is one of those flowsheets that can be simple or impossible depending on the answer to one question: what are you actually selling? The USGS Minerals Yearbook defines acid grade as fluorspar with a minimum CaF2 content of 97%. That product goes to hydrofluoric acid manufacture. Metallurgical grade is everything below 97% CaF2, used mainly as a steelmaking flux. Cement grade is generally 40% to 50% CaF2, aimed at the cement industry. Ceramic grade sits in its own contract window, and no universal figure appears in the sources we can cite. You'll have to ask the offtaker for that number.

You'll notice I've not put hard SiO2 and sulphur limits in that table. That's deliberate. The 97% split is a published definition. The actual silica and sulphur maxima are written into each offtake contract and an HF acid plant's feed specification, not into any universal standard we can cite here. If the buyer cannot produce that written spec, you don't yet have a project. You have an idea for a project.

The USGS 2026 summary notes the United States has been 100% net import reliant for fluorspar in 2025, excluding sales from stockpiled fluorspar, and significant mine production has not been reported since 1995 (Mineral Commodity Summaries 2026). That import reliance doesn't change your flowsheet, but it changes the offtake conversation. A buyer in a net import reliant market may pay more attention to acid grade than metallurgical grade, because acid grade is the feedstock for a strategic chemical.

You'll find similar specification-first thinking on our industrial mineral solutions pages.

Calcite and barite share the same fatty acid collector chemistry as fluorite. That's the core problem. You cannot simply float fluorite and leave them in the tailings, because they want to float too. Selectivity has to come from depressants, pH and water chemistry, not from a magic collector. Sodium silicate, tannin extracts, lignosulphonates and dextrin are the common directions. Each has a dosage window. Push the dosage too far and you'll depress fluorite as well. Under-dose and you'll get a concentrate full of carbonate.

For barite type ore, mixed flotation followed by separation is the usual route. Float all three minerals together off the gangue, then separate fluorite from barite and calcite in a second stage. This is not a beginner's circuit. The separation often needs a different pH, a different depressant, and much tighter water control than the roughing stage. A common mistake is to run the rougher at a high pH and then wonder why selectivity collapses in the cleaner. The pH that works for roughing may not be the pH that works for separation. You'll often run two different pH setpoints in the same plant.

Acid grade fluorite requires repeated cleaning. Trace carbonate and silica that survive roughing will fail the 97% CaF2 specification. So you rough, then clean, then clean again. Each cleaner stage lifts the grade but drops recovery. That's the trade. Regrinding between cleaner stages is often needed to liberate locked carbonate and silica. If you don't regrind, you'll carry locked gangue into the final concentrate. If you over-grind, you'll make fines that hurt filtration and recovery.

The number of cleaner stages is an economic decision driven by the gap between head grade and the written specification. A high head grade with low carbonate might reach acid grade with fewer cleaner stages than a low grade, carbonate rich ore. There is no fixed number. The testwork determines how many stages are justified, and the offtake margin tells you whether that circuit still makes money. The economic decision is not simply how many cleaners. It's how much fluorite you are willing to lose to hit the grade. The gap between head grade and written specification tells you that. Testwork gives you the grade recovery curve for each cleaner stage.

Final concentrate moisture is a contract term for acid grade fluorspar, not an afterthought. An HF plant may penalise you for every percentage point above the specified moisture. Filtration performance depends on particle size, clay content and slurry temperature. These three variables must be tested on the actual concentrate, not on a hand-picked lab sample. Over-grinding in the cleaner circuit can create fines that slow filtration and raise final moisture. You can lose more money on moisture penalties than you saved by adding an extra cleaner stage. Test filtration early, with the real concentrate, at the expected solids density. Thickener underflow density, filter cloth selection and vacuum level all respond to the same variables. A clay rich concentrate will blind a filter cloth quickly. Test filtration with realistic clay content, because a clean lab sample will give you a misleadingly high filtration rate.

The same acid grade target that demands a long cleaner chain also demands dry, low moisture concentrate. That's not a contradiction. It's a single design constraint. If the offtake specifies a moisture limit, design the filter and drying circuit to hit it. If you don't know the moisture limit, you're not ready to size the thickener or the filter.