Froth Flotation: How the Process Works
Froth flotation is the workhorse of mineral processing—here's the chemistry, circuit logic, and sizing rules that determine whether a plant performs.
Discuss your project
Overview
Froth flotation is a separation process that uses air bubbles and chemical reagents to make valuable mineral particles hydrophobic so they attach to bubbles and float, while gangue stays behind. You'll find it at the heart of most sulfide and oxide concentrators. If you're specifying a new plant, understanding the surface chemistry and circuit arrangement early will save you from a design that underperforms.
A flotation circuit rarely works as a single bank. You run rougher, scavenger, and cleaner stages because each stage has a different job. Before we get to that, let's look at why a particle floats at all. You can see how this fits within a larger mineral processing flowsheet.
Surface chemistry, in plain terms
Froth flotation depends on wettability. A hydrophilic surface wets with water and stays in the pulp; a hydrophobic surface repels water and attaches to an air bubble. Most valuable minerals aren't naturally hydrophobic enough, so you add collectors. A collector is a reagent that adsorbs onto the target mineral surface and makes it hydrophobic. Frothers are another reagent class; they reduce bubble size and stabilise the froth layer. Modifiers—such as pH regulators, activators, and depressants—adjust the surface chemistry so one mineral floats while another doesn't.
Conditioning follows a definite order. First you adjust pH to a range where your collector adsorbs selectively. Then you add collector and give it time to coat the target particles. After that you add frother just before the pulp enters the cells, because frother controls bubble size and froth stability during flotation. The exact reagent suite depends on ore mineralogy and is confirmed by test work, not guesswork. The Canadian Institute of Mining, Metallurgy and Petroleum (CIM) treats flotation as a surface-chemistry separation, and its technical references explain the role of contact angle in bubble attachment.
The rougher-scavenger-cleaner circuit, and why it's arranged that way
A conventional flotation circuit has three jobs in sequence: rougher, scavenger, cleaner. The rougher stage takes the conditioned pulp first and aims for maximum recovery, accepting a low concentrate grade. The scavenger stage then treats the rougher tailings and tries to recover any remaining hydrophobic particles. The cleaner stage takes the rougher and scavenger concentrates and re-floats them, often several times, to raise the concentrate grade.
You arrange it this way because recovery and grade fight each other. If you try to make a high-grade concentrate in a single pass, you'll float too little mass and lose recovery. If you float everything in one rough pass, you'll pull in gangue and dilute the concentrate. The rougher-scavenger-cleaner layout lets you push recovery in the first two stages and grade in the last stage. Middlings—particles that are neither clean concentrate nor final tailings—usually recycle to a previous stage or a regrind mill. Scavenger concentrate typically returns to the rougher feed, while cleaner tailings may go to regrind and then back to the scavenger or cleaner feed. That recycle loop is a core reason the circuit works as a system, not as isolated cells.
Xinhai's test records show this arrangement in practice. For example, a Kazakhstan copper flotation test used a rougher, two scavenger, and three cleaner stages at a grind of 50% passing 200 mesh. According to the company's technical documentation, that test produced a 16.23% Cu concentrate at 66.23% recovery. Those numbers are specific to that ore, but the circuit logic is universal.
Flotation cell types and sizing
Mechanical flotation cells are the most common. They use a rotating impeller to disperse air and keep solids suspended. Column cells, in contrast, introduce air through spargers and rely on a deep froth zone for selective upgrading. Tank cells and pneumatic cells trade mixing intensity for lower energy or finer bubble control. You choose a cell type based on ore kinetics, froth stability, and the stage's job. Roughers often use large mechanical cells for high throughput; cleaners may use columns for better selectivity. A rougher bank usually has several cells in series; each cell removes a fraction of the floatable material. Scavenger banks are similar but treat depleted pulp, so their froth is lower grade. Cleaner banks are shorter because the concentrate mass is small and residence time needs are lower.
Sizing starts with residence time. You need enough cell volume for the slowest valuable particles to attach to bubbles. Froth area matters too, especially in cleaners, because the froth must carry concentrate before it collapses back into the pulp. Manufacturers publish cell volumes and air dispersion data. Xinhai reports it manufactures flotation equipment with single-cell volumes as large as 320 m³. That's an engineering limit, not a default; your ore's flotation kinetics drive the final number and arrangement.
Grind size and liberation
Liberation is the condition where valuable mineral grains are detached from gangue. You grind to a size where most target particles are free enough to float on their own. If the grind is too coarse, locked particles won't float and report to tailings. If it's too fine, you create slimes that consume reagents, reduce bubble attachment, and hurt selectivity. Liberation isn't binary; you can have partial liberation, where a particle contains both valuable and gangue minerals. Flotation may still recover it, but the concentrate grade suffers. Regrinding cleaner tails is a common way to improve liberation without overgrinding the whole feed.
The right grind is usually determined by metallurgical testwork that maps recovery against particle size. P80 is a common grind target. It's the screen opening at which 80% of the sample passes. In the Kazakhstan copper test mentioned above, the grind target was 50% passing 200 mesh, which is 74 micrometres. That's a mid-range grind for a porphyry-type copper ore, but it isn't a universal setting. Every ore has its own liberation size. Overgrinding is a common mistake because it wastes energy and can lower recovery.
Recovery versus concentrate grade: the trade-off
Recovery is the percentage of the valuable mineral or metal that reports to the concentrate. Concentrate grade is the percentage of valuable content in that concentrate. These two metrics pull in opposite directions. If you let more particles float, recovery rises but grade falls because gangue entrainment and accidental flotation increase. If you tighten conditions to reject more gangue, grade rises but you'll lose some valuable particles, so recovery drops.
That's why the rougher-scavenger-cleaner circuit exists. You measure the trade-off in test work and set operating conditions—collector dosage, frother addition, air rate, froth depth—to land on an economic optimum. A 66.23% recovery at 16.23% Cu grade, from the plant design and engineering stage, is a data point, not a target. The optimum depends on metal price, smelter terms, and downstream costs. You'll usually optimise revenue per tonne, not recovery alone.
Testwork, scale-up, and control
You can't design a flotation circuit from a textbook. You start with batch flotation tests on drill core or composite samples. Those tests produce a grade-recovery curve, which plots concentrate grade against metal recovery for different reagent dosages, grind sizes, and pH values. A grade-recovery curve is the single most useful output from ore dressing test work. It shows the economic frontier, not a single point.
From batch tests you move to locked-cycle tests, which simulate the recycle of middlings in a continuous circuit. After that, a pilot plant test may run a few hundred kilograms per hour to confirm kinetics, reagent consumption, and froth handling at scale. Scale-up then selects the number of cells and their arrangement by matching laboratory retention time and air dispersion. Operators control the circuit with level sensors, air flow meters, pH probes, and online grade analysers. The goal is to hold the circuit at the steep part of the grade-recovery curve, where a small change in recovery or grade gives the highest economic return.
Xinhai's pilot base supports mineral processing test programs before design. If you're evaluating a project, you'll want to start with a testwork discussion rather than a fixed flowsheet.
According to the USGS Mineral Commodity Summaries, flotation is the dominant separation method for sulfide copper, molybdenum, and lead-zinc ores. That's because the process can handle fine particles and complex mineralogy that gravity or magnetic methods can't match. When you evaluate a flowsheet, test work with your own ore is the only reliable basis for recovery and grade targets.
Frequently asked questions
What is froth flotation?
Froth flotation is a mineral separation method based on surface wettability differences. Valuable mineral particles are made hydrophobic with collectors so they attach to air bubbles and float to the surface, while hydrophilic gangue stays in the pulp.
Why are rougher, scavenger, and cleaner stages used in a flotation circuit?
Each stage has a different purpose. The rougher maximises recovery at low grade, the scavenger captures remaining values from rougher tailings, and the cleaner upgrades the combined concentrate to a saleable grade. This arrangement lets you push recovery and grade independently.
How does grind size affect flotation?
Grind size controls liberation. If particles are too coarse, valuable minerals remain locked inside gangue and won't float. If they are too fine, slimes form and reagent consumption rises while selectivity drops.
What is the difference between recovery and concentrate grade?
Recovery is the percentage of the valuable mineral or metal that reports to the concentrate. Concentrate grade is the proportion of valuable material by mass inside the concentrate. They pull in opposite directions, and the optimum is an economic trade-off.