From building blocks to plastics, fibres and detergents: Building blocks and their family trees

1.1 Ethylene goes mostly into polyethylene, with three other big branches

Ethylene is the largest building block by tonnage, and almost all of it is used for chemicals rather than fuel. Wittcoff sorts its chemistry into a few big jobs. It is polymerised directly into the three polyethylenes. It is oxidised to ethylene oxide (EO). It is chlorinated to ethylene dichloride (EDC) on the way to vinyl chloride and PVC. And it is joined to benzene to make ethylbenzene, the precursor of styrene. In the United States in 2000 the polyethylenes took about 53% of ethylene, EO 12%, EDC 11% and ethylbenzene 6%; together with alpha-olefins and vinyl acetate the listed uses took about 90%. Brennan puts polyethylene at around 60% of world ethylene demand.

Ethylene oxide is the branch where chemistry and money meet most clearly. Ethylene and oxygen pass over a silver catalyst at about 250 C and 15 bar. The same catalyst can also burn ethylene, and the EO itself, to carbon dioxide and water, a far more exothermic reaction. Selectivity (the share of reacted ethylene that becomes the wanted product) was 65-70% with early catalysts and averaged about 82% by 2004, with the newest catalysts above 90% at the start of a run.

In 2002 about 61% of US EO was hydrated to ethylene glycol. A huge excess of water, 18-24 moles per mole of EO, stops the glycol reacting on to di- and triethylene glycol, yet the three still come out at about 91:8.6:0.4 by weight. About 68% of glycol goes into polyester; most of the rest is antifreeze. The remaining EO becomes ethoxylate surfactants, ethanolamines and polyols, which return in Part 5.

• Wittcoff (2004), Ch 3 Chemicals and polymers from ethylene · pdf pp. 126, 152-154, 164-165
• Wittcoff (2004), Ch 2, Table 2.8 · pdf p. 99
• Brennan (2020), Ch 2 Market evaluation · pdf p. 55

1.2 Propylene's reactive methyl group gives it a wider spread of chemistry

For many years propylene was the by-product nobody wanted, sold at a little above fuel value or simply burned (Wittcoff). In the United States in 2001 about 49% came from steam crackers and 51% from refinery catalytic cracking, and a large share went into gasoline rather than chemicals. How much a cracker makes depends on its feed. Ethane gives almost none: the propylene to ethylene weight ratio is 0.019. Naphtha gives 0.43-0.57 depending on how hard it is cracked. As polypropylene grew, producers added on-purpose routes: propane dehydrogenation over platinum or chromia catalysts at 530-700 C, metathesis of ethylene with 2-butene, and deep catalytic cracking of heavy gas oil.

The chemistry differs from ethylene because of the methyl group next to the double bond. Its hydrogens are allylic, meaning easy to pull off. That lets propylene be oxidised at the methyl end to acrolein and acrylic acid, or ammoxidised with ammonia and air to acrylonitrile. The same feature spoils free-radical polymerisation, which is why useful polypropylene had to wait for Ziegler-Natta catalysts (Part 2).

In the United States in 2001 polypropylene took 51% of propylene, acrylonitrile 12%, propylene oxide 10.5% and cumene 10%. Acrylonitrile is made in a fluidised bed because the reaction releases so much heat; one catalyst gives full propylene conversion with 86.9% selectivity at 320 C, plus hydrogen cyanide as a by-product. Propylene oxide shows the co-product problem. The old chlorohydrin route wastes chlorine, while the hydroperoxide routes make about 2.2 moles of tert-butanol, or 2.2-2.5 lb of styrene, per unit of propylene oxide, so profits hinge on selling the second product.

• Wittcoff (2004), Ch 4 Chemicals and polymers from propylene · pdf pp. 193-197, 207-213, 221-224, 236-237

1.3 Butadiene and the aromatics feed rubber, nylon and polyester

The C4 cut from naphtha cracking is roughly 47% butadiene, 23% isobutene and 25% n-butenes. Butadiene is recovered by extractive distillation: a polar solvent such as acetonitrile, N-methylpyrrolidone or dimethylformamide holds back the other C4s so that butadiene can be distilled out. About 70% of butadiene goes into synthetic rubber. In the United States in 2001 SBR took 27%, polybutadiene 25%, styrene-butadiene latex 11%, and nitrile and chloroprene rubbers 3% each. Another 14% went to hexamethylenediamine for nylon 66, made by adding hydrogen cyanide to butadiene over a nickel phosphite catalyst, the cheapest of several routes. ABS took 5%.

Benzene comes mainly from catalytic reforming in the United States and from pyrolysis gasoline (a liquid by-product of steam cracking) in Europe. Reformate is only 10-15% benzene but 50-55% toluene and 30-35% xylenes. In 2001 US benzene went 47% to ethylbenzene for styrene, 27% to cumene, 13% to cyclohexane for nylon and 8% to nitrobenzene for aniline and MDI. Toluene, the least wanted, is mostly converted: dealkylated with hydrogen to benzene, or disproportionated, two toluenes to benzene plus xylenes. A shape-selective zeolite can make that xylene 95-99% para-xylene.

Xylenes are a separation puzzle, since the four C8 aromatics boil within 9 C of each other. Para-xylene is taken out by low-temperature crystallisation or by adsorption on molecular sieves in a simulated moving bed (UOP's Parex process). The unwanted isomers are isomerised back to an equilibrium mixture over ZSM-5 and recycled until little is left. Almost all para-xylene is oxidised to terephthalic acid for PET.

• Wittcoff (2004), Ch 5 Chemicals and polymers from the C4 stream · pdf pp. 249-260
• Wittcoff (2004), Ch 7 Benzene, Ch 8 Toluene, Ch 9 Xylenes · pdf pp. 293-296, 338-340, 346-348
• Brennan (2020), Ch 2 Market evaluation · pdf p. 55

See it in the table

Check your understanding

Why does a 10-point gain in ethylene oxide selectivity matter so much to producers?

What makes propylene chemically different from ethylene in ways industry exploits?

A country cracks mainly ethane. Which statement best follows from Brennan's figures?