Chlor-Alkali Process: Cells, Products and Mercury Phase-Out

Chlor-alkali is the industrial process that makes chlorine (Cl2) and sodium hydroxide (caustic soda, NaOH) together, along with hydrogen, by passing electricity through a salt solution. Three cell designs do this work (mercury, diaphragm and membrane), and they differ mainly in how they keep the chlorine apart from the caustic. Membrane cells are now the reference technology; mercury-cell production is being phased out.

Definition and scope

The European Commission’s reference document on best available techniques for the chlor-alkali industry (the BREF, published October 2000) defines the sector as the production of chlorine and sodium hydroxide or potassium hydroxide by electrolysis of a salt solution. Sodium chloride is the usual feed; potassium chloride is used less often. The same document states that 95% of world chlorine production was obtained by the chlor-alkali process at the time it was written, and that the chlorine output of a country is an indicator of the state of development of its chemical industry.

The overall reaction is the same in every cell: sodium chloride and water go in, and chlorine gas, hydrogen gas and a sodium hydroxide solution come out. At the anode, chloride ions are oxidised to chlorine (2 Cl⁻ → Cl2 + 2 e⁻). At the cathode, water is reduced to hydrogen and hydroxide ions. What distinguishes the three technologies is the method used to keep the chlorine formed at the anode away from the caustic soda and hydrogen formed, directly or indirectly, at the cathode.

How the three cell types work

Mercury cell (Castner-Kellner, 1892)

In the mercury process the cathode is flowing mercury. Sodium produced at the cathode dissolves in the mercury as a sodium amalgam, which flows to a separate decomposer where it reacts with water to give hydrogen and sodium hydroxide, and the mercury is returned to the electrolyser. Because the caustic forms in a separate vessel, the BREF notes that 50% caustic soda is obtained directly from the decomposers.

Diaphragm cell (Griesheim, 1885)

In the diaphragm process, purified brine percolates through a porous diaphragm from the anode compartment to the cathode compartment. The BREF explains the reason for the diaphragm: without it, hydrogen and chlorine would ignite, and sodium hydroxide and chlorine would react to form hypochlorite and chlorate. The liquor leaving the cell contains 10-12% NaOH and about 15% NaCl, so it goes to evaporators where solid salt and 50% caustic are recovered. Diaphragms have usually been made of asbestos, and the BREF identifies asbestos as the main issue with this technology, both for employee exposure and for releases to the environment. Non-asbestos diaphragms are also covered in the document.

Membrane cell (developed 1970)

In the membrane process a water-impermeable, ion-conducting membrane separates anode and cathode. Brine flows through the anode compartment where chloride ions are oxidised to chlorine; sodium ions migrate through the membrane to the cathode compartment, where water is hydrolysed to hydrogen and hydroxide ions. The caustic is typically brought to 32-35% by recirculation before it leaves the cell, and because the membrane blocks chloride, the caustic contains no salt of the kind found in diaphragm-cell liquor. Evaporation is needed only if 50% caustic is wanted. The BREF describes the membranes as two-layer perfluorinated materials reinforced with PTFE, with an economic life of roughly three years (range 2-5 years), and notes that the feed brine must be very pure. It also reports that since 1987 practically 100% of new chlor-alkali plants worldwide have used the membrane process, while replacing existing mercury and diaphragm capacity has been slow because of the long life of those plants and the capital cost of replacement.

The data table further down this page compares the three cell types side by side, using only values stated in the BREF.

Products and the fixed co-production ratio

Chlorine and caustic soda come out in a fixed proportion set by molecular weights (40.00 / 35.45 = 1.128). The BREF states that electrolysis yields 1.128 tonnes of caustic soda (as 100% NaOH) per tonne of chlorine. It calls the co-production of the two in almost equal amounts a long-standing problem for the industry, because they serve very different end uses with different market dynamics, and demand for the two coincides only by chance. The document names PVC (through vinyl chloride monomer) as the driver of chlor-alkali production in most European countries, and lists polyurethanes, chloroaromatics, propylene oxide, chlorinated solvents and inorganic chlorine compounds among other uses of chlorine. Hydrogen is the third product.

This fixed ratio is also why chlor-alkali is a standard example when life-cycle assessment practitioners discuss co-product allocation: one process step yields two saleable outputs, and the environmental burden has to be divided between them. CASRAI’s guide to life cycle inventory data sources and methods covers how allocation is handled. For industry classification, chlor-alkali sits within basic chemical manufacturing; see the dictionary entry on the NAICS code for chemical manufacturing.

Regulation and the mercury phase-out

Mercury is the thread running through the regulatory history, so the dates below are worth reading side by side. Each row is attributed to the document that states it.

Instrument	What it says about chlor-alkali	Source
EU BAT reference document (BREF), October 2000	Membrane technology is considered best available technique; non-asbestos diaphragm is also BAT. Energy benchmark for chlorine plus 50% caustic: below 3,000 kWh (AC) per tonne of chlorine with liquefaction excluded, and below 3,200 kWh (AC) per tonne with liquefaction and evaporation included.	European Commission IPPC BREF
Euro Chlor members’ commitment	European members committed to phase out mercury technology by the end of 2017. Under EU legislation, liquid mercury must be converted to mercury sulphide by the end of 2022. At the end of 2018, members reported 2,947 tonnes of liquid mercury on site, with 1,146 tonnes converted during 2018.	Euro Chlor, Chlor-alkali Industry Review 2018-2019
Minamata Convention, Annex B Part I	Lists mercury-cell chlor-alkali production among processes to be completely phased out by 2025, and requires parties to prohibit new facilities using mercury-based technologies and to facilitate the shift of existing ones.	Minamata Convention secretariat
US EPA, 40 CFR Part 63 Subpart IIIII	National emission standards for hazardous air pollutants for mercury cell chlor-alkali plants under Clean Air Act section 112(d); final rule 19 December 2003, with a residual risk and technology review proposed 8 January 2021 (as recorded on EPA’s archived January 2021 page).	US EPA

Two points about reading this table. First, these are separate instruments with separate scopes: an industry commitment, an EU legal requirement on stored mercury, an international treaty and a US air-emission rule. None of them substitutes for another. Second, the BREF’s regional capacity figures (for example, mercury cells at 55% of western European capacity in June 2000) are historical snapshots from 2000; Euro Chlor’s later review charts the shift from mercury to membrane over 2001-2018 but this page does not quote percentages from it.

The BREF also records the scale of the legacy issue: about 12,000 tonnes of mercury were contained in EU mercury cells in 2000, and it flagged that, when plants are converted or shut down, this mercury has the potential to be released unless it is managed. For general background on the element’s hazards, see why mercury is dangerous.

Where it matters for labs and chemical supply

Few laboratories run chlor-alkali cells, but many use their outputs: sodium hydroxide solutions, chlorine-derived disinfectants and reagents, and the plastics and solvents downstream of chlorine. Three points from the sources are useful context for lab and supply staff.

• Caustic strength depends on the process. Industrial caustic leaves a mercury cell at 50%, a membrane cell at 32-35% and a diaphragm cell at 10-12% with salt present (BREF), which is why salt content and concentration are properties a product specification can differ on depending on how the caustic was made and finished.
• Chlorine safety is singled out. The BREF calls out the special importance of safety in the production, handling and storage of chlorine. For how corrosive caustic solutions are classified and labelled, see corrosive fluids: what counts as corrosive, and for reading supplier hazard information, how to read a safety data sheet.
• Process choice is an environmental-compliance question for manufacturers, not a lab-bench one. The mercury, asbestos and energy issues in the table above affect plant permitting and reporting, not how a purchased reagent is used.

Common misunderstandings

Is chlor-alkali a type of chemical, or a process?

It is a process and an industry sector. The name combines “chlor” (chlorine) and “alkali” (the caustic soda or potash made alongside it). It does not name a single product.

Are all three cell types still in use?

Yes, historically all three have operated in parallel. The BREF reported the mix differing by region in 2000 (mercury dominant in western Europe, diaphragm in the United States, membrane in Japan). The direction since is toward membrane, driven by new-plant choices and by mercury policy,.

Can a plant make chlorine without caustic?

Not by this route. The BREF states the two are produced in fixed proportions, and that the demand mismatch between them has long been an industry problem.

Does membrane mean salt-free everywhere?

The membrane’s advantage is that chloride does not migrate to the cathode compartment, so the caustic contains no salt of the kind found in diaphragm-cell liquor. The brine fed to the cell, however, must be very pure.

Frequently asked questions

What does chlor-alkali mean?

It refers to the electrolytic production of chlorine and an alkali, usually sodium hydroxide, from a salt solution. Hydrogen is produced as well.

What are the three chlor-alkali processes?

Mercury cell, diaphragm cell and membrane cell. They share the same overall reaction and differ in how the anode and cathode products are kept apart.

How much caustic soda is produced per tonne of chlorine?

1.128 tonnes of caustic soda, counted as 100% NaOH, per tonne of chlorine, according to the 2000 EU BREF.

Why is mercury-cell chlor-alkali being phased out?

The Minamata Convention lists mercury-cell chlor-alkali production among processes to be completely phased out by 2025, and the EU BREF describes the membrane process as avoiding both mercury and asbestos while also being the most energy efficient.

Sources

• European Commission, Integrated Pollution Prevention and Control (IPPC), Reference Document on Best Available Techniques in the Chlor-Alkali Manufacturing Industry, October 2000. PDF. Read 2026-11-12.
• Euro Chlor, Chlor-alkali Industry Review 2018-2019. PDF. Read 2026-11-12.
• Minamata Convention on Mercury, Manufacturing processes in which mercury is used. Read 2026-11-12.
• US EPA, Mercury Cell Chloralkali Plants: National Emissions Standards (archived January 2021 snapshot). Read 2026-11-12.