Every cup of coffee leaves something behind, and the volume is staggering. According to the figures in a new study out of Spain, the world produces roughly 10 million metric tons of coffee beans a year, and a very large share of what is brewed from them ends up as wet, brown sludge in a bin. A research group at Universitat Rovira i Virgili (URV) in Catalonia has now described a plain, moderate-temperature way to pull the valuable oil out of that sludge, and the approach has drawn headlines as a "surprisingly simple" route from coffee waste to fuel.
The work appears in the journal Biomass and Bioenergy (2026, volume 213, article 109467) under the title "Systematic evaluation of batch hexane extraction as a scalable pretreatment for the comprehensive valorization of spent coffee grounds." The authors, all tied to URV's Department of Chemical Engineering, are Jorge F. Romero, Alberto Tampieri, Daniel Montané, Magdalena Constantí and Francesc Medina. ScienceDaily ran the story on November 14, 2026, using material supplied by the university.
Why Spent Grounds Are Worth a Second Look
Used grounds look like trash, but chemically they are not empty. The researchers note that spent coffee grounds contain roughly 15 percent lipids, meaning fats and oils that survive the brewing process. Those lipids are the raw material for biodiesel, the same family of fuel that is made from soybean, canola and recycled cooking oil. Instead of growing a crop for fuel, the idea is to harvest a fuel ingredient from something that has already served its purpose.
The coffee angle also has an environmental side. Coverage of the study points out that more than 6 million tonnes of spent grounds are generated every year and that much of it goes to landfill, where decomposing organic matter releases methane. Turning even part of that stream into a product changes the math for anyone who has to dispose of it, from a roaster to a hotel chain to a university cafeteria. Notably, the grounds used in this study came from a single university cafeteria, which says something about how easy the starting material is to find.
What the Spanish Team Actually Did
The experiment centered on a method called batch extraction with n-hexane, a common solvent for separating fats. The grounds were dried and sieved to particles under 500 micrometers, then mixed with hexane in a vessel. Rather than guess at settings, the team varied three things at once: the temperature, how long the mixture was processed, and how much solvent was used for each gram of dry grounds. According to one report on the paper, they worked through 27 combinations of those conditions, with every test repeated, for 54 observations.
They then compared the results against Soxhlet extraction, the standard laboratory benchmark for pulling oil out of solid material. Soxhlet is thorough but slow. In the reported comparison it took 24 hours and yielded 16.6 percent crude oil from the dry grounds. The new batch method took one hour.
The Settings That Won
The best conditions were 45 degrees Celsius for 60 minutes, using 35 milliliters of hexane per gram of dry residue. Constantí, the study's lead author, put it plainly: "the optimal conditions are at 45°C for 60 minutes with a ratio of 35 millilitres of hexane per gram of dry residue." Forty-five degrees Celsius is about 113 degrees Fahrenheit, roughly the temperature of a hot bath, and well below hexane's boiling point of 68.7 degrees Celsius, which is a figure given in the paper coverage. Temperature turned out to be the biggest lever on yield, and yield fell off as the mixture approached the boiling point. Recovery leveled out after about an hour, and adding more or less solvent beyond the sweet spot gave no benefit.
Oil Yield and Purity
At those settings the method recovered about 90 percent of the oil that Soxhlet extraction pulls out. That is not a record, and the researchers do not claim it is. What stands out is the quality. Impurities, meaning material that does not convert into the fatty acid methyl esters that make up biodiesel, came to 0.3 percent of the new extract, compared with 3.9 percent for the Soxhlet oil. In practical terms, the faster method gave up about a tenth of the oil in exchange for a much cleaner product.
The fatty acid makeup was dominated by linoleic and palmitic acids, and that profile stayed consistent across the different conditions. One report on the paper gives approximate shares of 44 percent linoleic, 35 percent palmitic, 8 percent oleic and 7 percent stearic acid. A steady, predictable composition matters for anyone hoping to turn the oil into fuel, because it suggests the output would not swing wildly from batch to batch.
Why Not Use Faster Gadgets?
An obvious question is whether newer technology could speed things up further. The team tested ultrasound-assisted and microwave-assisted extraction, two approaches often pitched as shortcuts. They did start faster. Microwave extraction reached 12.5 percent oil in five minutes and 14.7 percent in 20, and ultrasound reached 12.5 percent in 15 minutes and 14.3 percent in 30, according to the figures reported from the paper.
Even so, the authors concluded that neither delivered enough of an overall advantage. Coverage of the study says the extra equipment did not bring meaningful gains in oil quality, efficiency, energy demand or scalability. That finding is what gives the study its "simple" label. A warmed vessel and a common solvent is hardly exotic, and the researchers describe batch hexane extraction under moderate conditions as the more balanced choice for a biorefinery.
The Leftover Grounds Are the Other Half of the Story
Medina, one of the co-authors, stressed that the oil is not the whole prize. "extracting the oil does not mean that the rest of the material cannot be used for something else," he said. The researchers designed the process to leave the grounds' lignocellulosic structure intact. That structure is the tough fibrous framework of cellulose, hemicellulose and lignin that makes up the bulk of the plant material.
The reported composition of the dry grounds was 8.8 percent cellulose, 30.9 percent hemicellulose, 17.9 percent lignin and 6.2 percent ash. After extraction, the cellulose and hemicellulose were preserved and the lignin concentration was above 20 percent. The authors suggest that the defatted solids could then feed other processes, including bioethanol, lactic acid, polyhydroxyalkanoates (a family of biodegradable plastics) and lignin-derived precursors for sustainable aviation fuel or phenolic compounds. Removing the fats may also act as a pretreatment, giving solvents and catalysts better access to what remains.
Montané framed the larger goal as a way "to transform a typically underused waste product into various energy vectors and bio-based chemical products." In that picture, spent coffee is a feedstock for a small biorefinery that makes several products, not a one-trick source of diesel.
What the Study Does Not Show
The headline claim deserves a careful reading. This is an extraction study. The reporting available on the paper does not describe the team producing biodiesel from the oil or testing it in an engine. The case for biodiesel rests on the oil's fatty acid profile and its low impurity level, which the researchers say point to that potential. The downstream products from the leftover solids, such as bioethanol and lactic acid, are likewise listed as possibilities, not demonstrated results.
Scale is another open question. The work was done as a laboratory batch process on grounds from one cafeteria. The coverage gives no pilot-plant or industrial data, no cost analysis and no life-cycle assessment, even though the paper's title calls the method a "scalable" pretreatment. The coverage also does not discuss how hexane, a petroleum-derived and flammable solvent, would be handled, recovered and recycled at commercial volumes. Those are the details that usually decide whether a lab result becomes a business. Nor do the available reports describe what the team plans to do next.
It is also worth keeping the researchers' own numbers in proportion. Recovering about 90 percent of the benchmark yield means some oil stays behind, and the oil recovered is a share of the roughly 15 percent of the grounds that is fat to begin with. The prize is a modest slice of each kilogram, and the logic depends on volume.
A Crowded Field of Coffee Waste Ideas
The URV paper joins a long line of attempts to find a second life for coffee residue. Researchers at Lancaster University reported in 2017 on a method that combined oil extraction and biodiesel production into one step, using methanol and a catalyst, with an optimal time of about 10 minutes, based on coverage of that work. More recently, a team in South Korea described converting wet grounds into a coal-like biochar in about 90 seconds, a different route that burns the grounds instead of extracting their oil. Each approach makes a different trade between speed, equipment and what ends up in the product. The Spanish group's contribution is a systematic map of the settings that matter for the solvent route, with a clear comparison against both the lab standard and the gadget-heavy alternatives.
What It Means for a Coffee Drinker
For the person making a pot at home, nothing changes this week. No fuel station is selling a coffee-grounds blend, and the coverage of the study mentions no commercial partner. The relevance is further upstream. Coffee is a global commodity at a scale of about 10 million tonnes a year, and any process that can pull value from the leftovers makes the whole chain a little less wasteful. If a method like this proves workable, the first beneficiaries would likely be places where grounds pile up in bulk: roasters, large cafes, instant coffee plants and institutional kitchens.
For now, the study offers something modest but useful: a documented recipe for getting a clean oil from a cheap waste stream at gentle temperatures in an hour, with the rest of the material left in a form that can be put to work. Whether it can be run safely and cheaply at industrial volumes is the question the next round of research will have to answer.

Comments
Comments are moderated. Please keep it civil — see our community guidelines.
No comments yet — be the first.