At your local coffee roaster you’re likely to find a host of beans on offer—from acidic, juicy-tasting light roasts to bold, rich dark roasts. But compare a light roast from one roaster to that of another, and you might find that the two cups don’t have the same perceived strength.
A team of chemists hoping to put a number on roast strength has found that simple cyclic voltammetry can do the trick (Nat. Comms. 2026, DOI: 10.1038/s41467-026-71526-5).
Christopher Hendon, a computational and food chemist at the University of Oregon who led the work, says that until now “there has basically been no way to assign the quality of coffee with a numerical value in any capacity. . . . If you wanted to say something tasted good, you’d rely on a human to taste it.”
Cyclic voltammetry can identify beans of different roast strength, like the ones above. Credit:
Shutterstock
Hendon says other analytical methods, like chromatography and mass spectrometry, can identify what compounds are in coffee but don’t necessarily correlate that to flavor. He says the flavor of coffee is mostly determined by the amount of total dissolved solids in a cup—which varies by brew method—and by how long the coffee was roasted. Cyclic voltammetry is capable of determining both.
Hendon and his team brewed various coffees using the Specialty Coffee Association of America cupping protocols (PDF), then immersed several electrodes of a standard cyclic voltammetry setup into the coffee. After running the voltage sweeps, “there were these features that showed up in the reducing region, which turned out to be related to the acid splitting around in the coffee,” he says. “This acidic feature that shows up in the reducing region maps to both roast and intensity. So the feature grows if [the coffee is] more concentrated, but separately it is at lower intensity if the coffee is darker. And that’s because when you roast coffee dark you start to destroy these organic acids.”
The team was then able to demonstrate that the method did well to discriminate between coffees the same way a taster would. The team asked a roaster to send four samples, one of which the roaster had rejected for its taste. Using cyclic voltammetry, the team was able to correctly identify which of the coffees had been rejected. “We didn’t reject it with our tongue,” Hendon says. “We rejected it using E-chem.”
Max Barnhart is an assistant editor and life sciences reporter at C&EN.
2026 American Chemical Society
