Heavy inorganics: salt, sulphur, nitrogen and phosphorus: Salt: chlor-alkali, hydrochloric acid and soda ash

2.1 Brine electrolysis makes a fixed bundle: a tonne of chlorine, 1.1 tonnes of caustic soda and some hydrogen

Salt is mined as rock, evaporated from sea water, or dissolved underground and pumped up as brine. Kent's 2010 US prices run from about $8 a tonne for brine to $170 for vacuum-pan salt.

A chlor-alkali cell passes direct current through nearly saturated, purified brine. At the anode (the positive electrode), chloride ions give up electrons and form chlorine gas: 2Cl- → Cl2 + 2e. At the cathode (the negative electrode), water takes up electrons to form hydrogen and hydroxide ions, which pair with sodium as caustic soda. Overall: NaCl + H₂O → NaOH + 1/2 H2 + 1/2 Cl2. The products must be kept apart, or they react to form hypochlorite (bleach).

The industry therefore sells a fixed bundle, often called the electrochemical unit. In theory each kilogram of chlorine comes with about 1.13 kg of NaOH. Dryden gives 1.15 t of caustic and 283 Nm3 (26 kg) of hydrogen per tonne of chlorine. The world made 42.3 Mt of chlorine and 45.0 Mt of caustic soda in 2008 (Kent). Kent says it uses about 2% of the electricity generated in much of the industrial world.

The energy follows from physics. The minimum cell voltage is 2.31 V (Shreve). A modern membrane cell runs at about 2.96 V at 6 kA/m2: 2.26 V to split the salt, 0.13 V of electrode losses and 0.57 V of resistance losses (Kent). A tonne of chlorine is 14,100 moles of Cl2, which needs 28,200 moles of electrons. At 96,500 coulombs each, that is 2.72 billion coulombs. At 2.96 V this is 8.1 GJ, or about 2,240 kWh; at 96% current efficiency, about 2,330 kWh. Older cells at 3.4-4.3 V approach Dryden's 2,900 kWh per tonne, close to Shreve's 10.6 MJ per kg.

• Kent (2012), Ch 26 Salt, Chlor-Alkali, and Related Heavy Chemicals · pdf pp. 1021-1023, 1031-1035
• Shreve (1984), Ch 13 Chlor-alkali Industries and Ch 14 Electrolytic Industries · pdf pp. 249-250, 261
• Dryden (1997), Ch IIH Chlor-Alkali Industry · pdf pp. 206, 210-212

2.2 Mercury, diaphragm and membrane cells trade electricity, steam and purity against each other

The three cell types differ in how they keep chlorine and caustic apart. A mercury cell uses a flowing mercury cathode. Sodium dissolves in it as a dilute amalgam, which then reacts with pure water in a separate vessel to give 50% caustic directly, with salt as low as 30 ppm (Shreve). A diaphragm cell separates the electrodes with a porous mat, historically asbestos. Only about half the chloride is converted, and the cell liquor is about 10-12% NaOH with about 15% NaCl. A membrane cell uses a perfluorinated ion-exchange sheet that passes sodium ions and water but blocks chloride, giving 30-35% caustic with chloride in tens of ppm (Kent).

Mercury cells use 10-15% more electricity than diaphragm cells but need no evaporation (Dryden). Diaphragm liquor needs multiple-effect evaporation: Shreve says about 2,600 kg of water must be boiled off per tonne of 50% caustic, against 715 kg from a membrane cell, and the product keeps about 1% salt. Membrane cells use about 77% of the electricity of mercury cells (Dryden), but need brine purified to parts per billion, and the membranes are costly.

Regulation and energy prices settled the contest. Mercury released from cells turned into methyl mercury in fish, and Japan ordered its mercury cells replaced by 1984 (Shreve). Kent calls the membrane cell the universal choice for new plants, yet in 2008, 67% of US capacity was diaphragm. Gulf Coast producers had cheap brine and cheap fuel for evaporation, which suited cogeneration. Membrane cells passed half of European capacity in 2011.

India shows the opposite pattern. In 1989 mercury cells made 88% of Indian caustic, diaphragm cells 7% and membrane cells 4% (Dryden). Conversion was slowed by equipment cost, technology fees and a 35% import duty.

• Kent (2012), Ch 26 Salt, Chlor-Alkali, and Related Heavy Chemicals · pdf pp. 1032-1037
• Shreve (1984), Ch 13 Chlor-alkali Industries and Ch 14 Electrolytic Industries · pdf pp. 250-254
• Dryden (1997), Ch IIH Chlor-Alkali Industry · pdf pp. 206, 212-217

2.3 Chlorine and caustic sell into different markets, so one of them is always the problem

The fixed ratio means a producer cannot make more chlorine without more caustic. Whichever product is in surplus must be sold cheaply, stored or destroyed, which keeps both markets volatile (Shreve).

The two markets have little in common. Chlorine is a building block for plastics and organic chemicals. In the US, ethylene dichloride (EDC) and vinyl chloride for PVC took 43% of chlorine and other organics 29% (Kent). Caustic soda goes into organic and inorganic chemicals, pulp and paper, alumina refining, soaps and textiles. Chlorine is hazardous to ship, so it is used on site, piped, or liquefied. Caustic travels easily as 50% solution. In the US, growth has come through large plants integrated forward into EDC.

Which side is short depends on the economy. Dryden records that in India in 1964 only about 65% of the chlorine made was used, with the rest burned to HCl or destroyed. He saw new outlets, polyvinyl chloride and chlorinated hydrocarbons, as the way to balance the two. In developed countries the problem ran the other way.

Hydrochloric acid ties the system together. Over 93% of US hydrogen chloride is a by-product: of chlorinating hydrocarbons, of isocyanate and fluorocarbon production, and of cracking EDC to vinyl chloride (Kent). Oxychlorination closes the vinyl loop by reacting the HCl with oxygen and ethylene, so the chlorine ends up in the product. Most synthetic acid is made by burning hydrogen in chlorine, often the tail gas left after liquefaction, which puts two co-products to use at once.

• Kent (2012), Ch 26 Salt, Chlor-Alkali, and Related Heavy Chemicals · pdf pp. 1031, 1036-1038
• Shreve (1984), Ch 13 Chlor-alkali Industries and Ch 14 Electrolytic Industries · pdf pp. 245-247, 254-255
• Dryden (1997), Ch IIH Chlor-Alkali Industry · pdf pp. 207-209, 216

2.4 Soda ash: Solvay recycles ammonia but cannot escape calcium chloride, and trona beats it on energy

Soda ash (sodium carbonate) is the cheapest soluble alkali. Glass took 46% of US use in 2010 (Kent); in India, soaps and detergents took 40% (Dryden).

The Solvay process aims at a simple overall reaction, 2NaCl + CaCO₃ → Na₂CO₃ + CaCl₂, which will not go directly. Ammonia makes it work. A kiln burns limestone to lime and a gas of over 40% CO₂. Purified brine absorbs ammonia, then flows down carbonating towers against CO₂. Sodium bicarbonate, sparingly soluble in this liquor, crystallises out at about 27 C. Filtered and heated, 2NaHCO3 → Na₂CO₃ + H₂O + CO₂, it gives light soda ash and returns half the CO₂. The filtrate holds ammonium chloride, which is boiled with milk of lime to free the ammonia for reuse, leaving calcium chloride (Kent).

Ammonia is a costly inventory, so recovery exceeds 99.8%, with losses of about 1-1.5 kg per tonne of product (Dryden). Salt is used less well. Dryden's 1.55 t of salt per tonne compares with a theoretical 2 x 58.5 / 106 = 1.10 t, about 71% use, close to his 75% bicarbonate yield. The stoichiometry also leaves about 111 / 106 = 1.05 t of calcium chloride per tonne of soda ash. It is usually discharged as dilute liquor into rivers (Kent).

Natural soda ash avoids most of this. Trona, a hydrated sodium carbonate-bicarbonate mined at Green River, Wyoming, only needs dissolving, purifying, crystallising and calcining. About 1.8 t of trona gives 1 t of soda ash. Shreve gives 18.4 GJ per tonne for Solvay ash against 8.4 GJ for natural ash. The last US Solvay plant closed in 1986. Trona is scarce elsewhere, so in 2010 the world still made 34.3 Mt of synthetic ash against 11.7 Mt of natural ash (Kent).

• Kent (2012), Ch 26 Salt, Chlor-Alkali, and Related Heavy Chemicals · pdf pp. 1023-1025
• Shreve (1984), Ch 13 Chlor-alkali Industries and Ch 14 Electrolytic Industries · pdf pp. 243-248
• Dryden (1997), Ch IIH Chlor-Alkali Industry · pdf pp. 197-205

See it in the table

Check your understanding

Using a membrane cell at 2.96 V and full current efficiency, about how much electricity does a tonne of chlorine need?

Why did diaphragm cells survive longest on the US Gulf Coast?

What is the main waste from the Solvay process, and roughly how much per tonne of soda ash?