Biorefineries and platform chemicals: Platform chemicals: the DOE top twelve and their families

4.1 How the DOE narrowed 300 candidates to twelve

PNNL and NREL started with over 300 possible building blocks compiled from earlier studies. Ranking them on feedstock cost, processing cost, market size and price did not separate the sugar-based candidates well, so the team switched to the petrochemical model: all petrochemical products descend from 8 or 9 foundation chemicals, so look for biomass molecules that could found families.

The filters were explicit. A good building block has several functional groups, so many derivatives can be made from it. It should be producible from both starch and lignocellulose, have one to six carbons, not be a lignin aromatic, and not already be a 'super commodity'. That last rule excluded ethanol and ethylene glycol, which compete against huge, depreciated petrochemical plants. No two-carbon molecule survived. Syngas products were set aside except hydrogen and methanol.

The team also sorted opportunities by strategic fit. Direct replacements, such as acrylic acid from lactic acid instead of propylene, enter known markets but compete purely on cost against depreciated capital. Novel products, such as polylactic acid, have no petrochemical rival but uncertain markets and long development times. Building blocks combine both and let a plant swing between derivatives.

About 50 candidates became a top 30, scored on strategic fit, value, technical complexity of both steps (sugar to building block, building block to derivative) and family potential. Twelve scored above average. Lactic acid, levoglucosan and lysine sat exactly at the mean and joined a second tier with gluconic, malonic, propionic, citric, aconitic and xylonic acids, acetoin, furfural, serine and threonine.

• Werpy & Petersen (2004), Executive summary and Sections 3-5 · pdf pp. 8-9, 12-23
• Werpy & Petersen (2004), Section 8 Moving forward · pdf p. 27
• Kamm (2006), Vol 1 Ch 1 Biorefinery systems - an overview · pdf p. 54

4.2 The glucose platform: one sugar, many families

Glucose is the hub. Kamm and colleagues call it predestined for a key position because both microbes and chemists can turn it into so much. Their fermentation family tree runs from one-carbon products (methane, methanol) through ethanol and acetic acid, lactic acid, propanediol and acetone, to butanol, itaconic acid, citric acid and beyond.

Three derivative families matter most commercially. The first replaces C4 petrochemicals. Succinic acid hydrogenates to gamma-butyrolactone, 1,4-butanediol and tetrahydrofuran, the same family made from maleic anhydride from butane. Task 42 puts the butanediol market near 1 million tonnes, about 40% of it going to THF.

The second replaces C3 petrochemicals. Lactic acid gives polylactic acid, ethyl lactate solvent, acrylic acid and propylene glycol. 3-Hydroxypropionic acid dehydrates to acrylic acid, a 3.3 million tonne market in 2006 per Task 42, and reduces to 1,3-propanediol.

The third makes aromatics-like rings without benzene. Dehydrating glucose gives HMF; oxidising HMF gives FDCA, proposed as a substitute for terephthalic acid in polyester. Hydrating HMF instead gives levulinic acid, whose derivative diphenolic acid could replace bisphenol A in polycarbonate.

Thoen and Busch of Dow add that glucose also yields sorbitol, whose hydrogenolysis gives glycols, and that the best chemistries play to biomass rather than force carbohydrates into a hydrocarbon pattern. Werpy notes many transformations remain unproven: selective reduction of acids in fermentation broth, robust catalysts and low-pH fermentation are the recurring gaps. A tree on paper is therefore a menu of options, not a set of working plants: of the 63 sugar-to-building-block pathways in Werpy's own table, only 14 were commercial.

• Kamm (2006), Vol 1 Ch 1 Biorefinery systems - an overview · pdf pp. 49, 53
• Kamm (2006), Vol 2 Ch 12 Industrial chemicals from biomass (Thoen & Busch) · pdf pp. 813-816, 818
• Task 42 (2012), Sections 9.3, 9.4 and 9.6 · pdf pp. 22-26
• Werpy & Petersen (2004), Section 7 Table 8 and Sections 9.1-9.8 · pdf pp. 26, 29-55
• Kamm (2006), Vol 1 Ch 7 The Biofine process · pdf p. 176
• Kent (2012), Ch 33 Biomass conversion · pdf p. 1291

4.3 The glycerol platform: a by-product looking for a market

Glycerol is the odd member of the twelve: it is not made from sugar but falls out of fats and oils. In 2004 Werpy and Petersen put world output at 500,000 to 750,000 tonnes a year, priced at 0.60 to 0.90 dollar per pound, which kept its markets small and fragmented. About 75% of US supply was natural, 25% synthetic from propylene.

Their key insight was that biodiesel would change the price. If the US displaced just 2% of on-road diesel with biodiesel by 2012, almost 800 million pounds of new glycerol would appear. They saw a floor of about 0.05 dollar per pound for crude glycerol, set by its value as feed or for steam reforming to hydrogen, and judged that at 0.20 to 0.50 dollar per pound it would become a major building block.

Their chemistry follows three families. Hydrogenolysis, breaking specific carbon-oxygen bonds, gives propylene glycol and 1,3-propanediol; the barrier is catalysts that tell C-C from C-O bonds. Oxidation gives glyceric acid and other branched polyester and nylon monomers. Direct polymerisation gives branched polyesters and polyols for polyurethanes.

Industry moved fast once biodiesel arrived. Task 42 reports ADM's 100,000 tonne propylene glycol plant on soy or canola glycerol, with an 80% cut in greenhouse gas emissions claimed. Solvay's Epicerol process converts glycerol to epichlorohydrin, the epoxy resin intermediate, with a 10,000 t demonstration plant at Tavaux since 2007 and a 100,000 t plant planned in Thailand. BioMCN makes methanol from glycerol, the main side-product of biodiesel, by way of syngas.

• Werpy & Petersen (2004), Section 9.10 Glycerol · pdf pp. 59-64
• Task 42 (2012), Sections 9.3 Propylene glycol and epichlorohydrin · pdf pp. 20-21
• Kamm (2006), Vol 1 Ch 17 Biocatalytic and catalytic routes (Willke et al.) · pdf pp. 418-421

See it in the table

Check your understanding

Why did the DOE screening exclude ethanol and ethylene glycol from its building block list?

Which pairing of building block and petrochemical target is correct?

What did Werpy and Petersen expect to turn glycerol into a major building block?