Bio-Based Polyurethane Foam: What Lignin and Castor Oil Actually Do to Rigid Foam

The case for bio-based foam is usually made as a straight substitution: replace a share of the petrochemical polyol with something that grew, keep the rest of the recipe, and bank the carbon saving. A study published in Scientific Reports in June 2024 makes that arithmetic look naive. Kraft lignin and castor oil, two cheap biomass-derived polyols, push rigid polyurethane foam in opposite directions, and the size of the divergence decides which products either material can realistically enter.

Hyeon Jeong Kim, Xuanjun Jin and Joon Weon Choi, working at Seoul National University's Graduate School of International Agricultural Technology in Pyeongchang, foamed a series of rigid polyurethanes with lignin at 5, 10, 15 and 20 per cent by weight of the polyol, and with castor oil from 10 to 100 per cent. Every batch had a control made under identical conditions with no bio-polyol at all. Lignin raised compressive strength as its loading rose, from 147 kPa at 5 per cent to 207 kPa at 20 per cent. Castor oil did the reverse: 284 kPa at 10 per cent, 23 kPa at 100 per cent. The control sat at 326 kPa.

For anyone who has tried to sell bio-content into construction, those numbers define the problem. A foam at 207 kPa is a packaging or void-fill product. A foam at 326 kPa is heading toward insulation board.

How a two-part foam recipe absorbs biomass

Rigid polyurethane is made from two streams that meet at a mixing head. The A-side is an isocyanate — in this case a polymeric methylene diphenyl diisocyanate supplied by BASF Korea as Lupranate M5S, with an NCO content of 32 per cent. The B-side is a polyol, here polyethylene glycol 400, plus a tin catalyst, a silicone surfactant, glycerol and a blowing agent. Hydroxyl groups on the polyol attack isocyanate groups and form urethane links; the blowing agent expands the reacting mass into cells that then set.

What makes a biomass feedstock usable as a polyol is hydroxyl groups — enough of them, in the right places, with the right accessibility. The researchers measured the lignin before they foamed it. It carried 4.9 millimoles of free hydroxyl per gram, split roughly 62 per cent phenolic, 33 per cent aliphatic and 5 per cent carboxylic acid, with a hydroxyl number of 328 milligrams of potassium hydroxide per gram. That is enough reactive sites to take part, and Fourier transform infrared spectroscopy confirmed urethane linkages formed in every bio-based batch.

Enough, though, is not the same as fast.

The index experiment that didn't behave

Polyurethane formulations run at an index: the molar ratio of isocyanate groups to hydroxyl groups. An index of 1.01 is close to stoichiometric. An index of 1.3 leaves a third more isocyanate than the hydroxyls can consume, and the excess reacts with itself and with existing urethane groups to add crosslinks. The team ran both, labelling the near-stoichiometric lignin foams LPU and the high-index versions lpu, with a matched control for each.

Read the table by comparison rather than by row. Pushing the control from index 1.01 to 1.3 added 115 kPa of strength. The same move bought the lignin foams almost nothing: 164 kPa against 147 kPa at the low loading, and a loss of 60 kPa at the high one. Extra isocyanate, the standard lever for stiffening a formulation, stops working once the polyol is lignin. That is the sort of finding a plant chemist needs before quoting a reformulation, and it is not visible in a headline figure comparing bio-foam to conventional foam.

Photo by Nick Fewings on Unsplash

Reactivity, not loading, sets the cell structure

The paper's explanation sits in the reaction kinetics. Both bio-polyols reacted more slowly with isocyanate than the polyethylene glycol they replaced, and the spectroscopy shows it. Lignin's hydroxyls are largely phenolic and sit on a bulky, polydisperse aromatic polymer, which slows the approach of isocyanate molecules. Castor oil's reactive group is a secondary hydroxyl on ricinoleic acid, the fatty acid that makes up about 90 per cent of the oil, and secondary hydroxyls are inherently slower than primary ones.

Slower reaction shifts the race between gelation and gas generation. Gel early against the blowing agent and the cells stay small and dense; gel late and the cells grow, thin and rupture. That is exactly what the scanning electron microscopy shows. Lignin produced smaller cells, with the smallest and most uniform structure at 10 per cent loading in both index series. Castor oil produced larger cells and a rising void fraction, tilting toward an open-cell structure as substitution climbed toward 100 per cent — an open-cell foam is a sponge, not an insulator.

Thermal behaviour divides along the same line. Char yield, the residue left after the foam decomposes in nitrogen, rose with lignin content from 11.5 to 15.8 per cent in the LPU series and from 12.4 to 17.5 per cent in the high-index series. With castor oil it fell from 10.6 per cent at 10 per cent loading to 4.2 per cent at full substitution. Char is what stands between a foam and a flame, and lignin builds it.

Where the feedstock actually comes from

Lignin is 25 to 30 per cent of lignocellulosic biomass, and the biosphere holds more than 300 billion tonnes of it. Pulp mills already separate roughly 50 million tonnes a year, then burn almost all of it for process heat: by common estimates about 2 per cent is recovered as a commercial product. Commercial kraft lignin capacity was estimated at roughly 112,000 tonnes in 2023 against a technically recoverable potential near 27 million tonnes at a 40 per cent mill outtake, according to analysis published in ACS Polymers Au. One frequently cited estimate puts recovered lignin burned for energy at around USD 0.20 per kilogram and the same material converted to chemicals near USD 1 per kilogram, which is the gap that funds separation plants.

The lignin in this study was softwood kraft lignin made by the LignoBoost process and supplied by Domtar, the paper company whose separation plant in Plymouth, North Carolina, was profiled in Chemical & Engineering News alongside Stora Enso's 50,000-tonne Finnish plant and West Fraser's 10,000-tonne facility in Alberta. Quality varies by wood species and by mill, a point the authors handle by characterising elemental content, ash and hydroxyl distribution before foaming. A formulator buying lignin by the truckload does not get that luxury for free.

Castor oil carries a different kind of concentration risk. Roughly 80 per cent of the world's castor seed and around 85 per cent of the traded oil come from India, with Gujarat state producing the bulk of it. Ricinoleic acid gives the oil a hydroxyl group that most vegetable oils lack, and because castor oil is inedible, industrial demand doesn't compete with food. But the crop depends on monsoon acreage, and a bad season in one Indian state moves the price for every foam formulator on three continents within a quarter or two.

The half of the recipe that stays petrochemical

A bio-polyol substitutes for the B-side. The A-side does not move, and in the European Union that A-side now carries a training obligation. Under REACH Annex XVII entry 74, introduced by Commission Regulation (EU) 2020/1149, diisocyanates cannot be used for industrial or professional purposes after 24 August 2023 unless the concentration stays below 0.1 per cent by weight or every user has completed training on safe use before handling them. Employers must document completion and renew it at least every five years; suppliers must label packaging and provide the training materials.

The requirement reaches the tasks where exposure actually happens: open mixing at ambient temperature including foam tunnels, spraying in a ventilated booth, cutting of articles that are not fully cured, cleaning and waste. The German Federal Institute for Occupational Safety and Health maintains a summary of the restriction and the training levels. Compliance costs scale with headcount, not with revenue, which means a small spray-foam contractor absorbs a heavier burden than a multinational panel maker. That asymmetry is part of the real price of any bio-polyol switch, and it isn't in the formulation data.

Photo by hayleigh b on Unsplash

Flame retardancy is the gate for building products

The same group published a follow-up in Materials Today Communications in 2025 that answers the next question, which is what happens when you need the foam to resist ignition. They reacted lignin with phosphorus pentoxide under reflux to make phosphorylated lignin, confirming 2.9 per cent phosphorus content by inductively coupled plasma analysis, then added it at 5 to 20 per cent by weight to foams already containing 10 per cent lignin or 40 per cent castor oil.

The mechanical result is the striking one. In the lignin foam, compressive strength climbed from 147 kPa to as much as 592 kPa, above the 326 kPa control. In the castor oil foam it rose from 207 kPa to 341 kPa. Added to a conventional foam, the same additive did the opposite and dropped firmness from 326 kPa to 260 kPa. Flame retardancy improved by 42 to 65 per cent, though not in proportion to how much additive went in. Ammonium polyphosphate, the commercial benchmark, delivered higher firmness and thermal stability overall while the phosphorylated lignin was comparable or better on some measures. The 2025 paper is the more commercially interesting of the two, because a foam that fails a fire test has no route into a wall.

What the second year of this work has to answer

Nobody has yet published the number buyers ask for first: thermal conductivity. The 2024 study reports compressive strength, apparent density, cell morphology, spectroscopy and thermogravimetric behaviour. It does not report lambda, water uptake, long-term creep under sustained load, or dimensional stability across a service life measured in decades. Rigid foam in a wall carries load at varying humidity for twenty years and more. Char yield is a proxy for fire behaviour, not a fire test result, which is precisely why the phosphorylated lignin follow-up matters more than its citation count suggests.

Then there is the reformulation bill. A polyol that reacts more slowly needs more catalyst, more isocyanate, or more oven time, and each of those eats into the bio-content gain and the cost saving that justified the change. Somebody at a plant has to run the trials, qualify the foam against a standard, and persuade a customer that a formulation with an agricultural supply chain behind it will arrive next quarter and the quarter after. That is a procurement problem as much as a chemistry one.

The study does offer one practical landmark. At 10 per cent lignin, both index series produced their smallest and most uniform cells, and the high-index variant held 167 kPa at that loading before falling away at 15 and 20 per cent. The peak compressive strength sits at 20 per cent. The best cell structure sits at 10. Anybody trying to hold a product specification will have to decide which of those two numbers their application actually needs, and the answer will differ between a cold-store panel and a cushioning insert.

That is the question waiting in year two: whether a formulation can be locked at the loading where the foam is most uniform rather than where it is strongest, and whether the lignin arriving at the silo is consistent enough, month after month, to keep it there. The full study and its open-access text give the starting conditions. The mill certificate gives the rest.