A new electrochemical method promises to do what refractometers and taste-testers alone cannot: measure both strength and roast quality in a single sip.
Dubai – Qahwa World
There is a quiet frustration that haunts every coffee lover’s life. You find a bag of beans you love. Bright, complex, perfectly balanced. You brew it exactly the same way the next morning. And somehow, it is wrong. Too bitter. Too sour. Thin and lifeless.
The problem is not your technique. Or rather, it is not only your technique. The problem is that coffee is one of the most chemically complex beverages on Earth. More than a thousand compounds interact in ways that scientists are still struggling to understand. And for decades, we have been flying blind when it comes to measuring what actually ends up in the cup.
The coffee industry has relied on a single number to assess quality: total dissolved solids, or TDS, measured by shining light through the liquid. A refractometer tells you how much coffee material is dissolved in the water. But it cannot tell you what that material is.
And that, as it turns out, is a serious problem.
Now, a team of chemists at the University of Oregon, led by Christopher Hendon, has published a study in Nature Communications that offers a radical alternative. They have shown that by running a simple electrical test on a cup of black coffee, with no sample preparation, no dilution, no fancy reagents, you can measure both the strength of the brew and, separately, how dark the coffee was roasted. Two of the most critical variables in coffee quality, captured in a single voltammogram.
The Refractometer’s Blind Spot
To understand why this matters, you have to understand what the coffee industry has been working with.
The refractometer is a marvel of practical engineering. It measures how much light bends as it passes through a liquid, the refractive index, and uses an empirical formula to convert that number into a percentage of dissolved solids. A typical filter coffee might register around 1.35% TDS, meaning that 98.65% of what is in your cup is water.
But here is the catch: different substances bend light differently. A 2% glucose solution has the same refractive index as a 4% ethanol solution. In a simple system, that is a problem. In coffee, which contains hundreds of organic acids, sugars, alkaloids, lipids, and melanoidins, it is a fundamental limitation.
Two coffees can have identical TDS readings and taste completely different. A light roast and a dark roast, brewed to the same strength, will produce wildly different flavor experiences. The refractometer cannot tell them apart.
Hendon’s team set out to build a tool that could.
A Method Borrowed from Battery Science
Cyclic voltammetry sounds intimidating, and the instruments used to perform it, potentiostats, are normally found in laboratories testing batteries or fuel cells. But the basic principle is elegant. You immerse electrodes in a solution, sweep the voltage across a range, and measure how much current flows.
Different molecules respond at different voltages, either donating or accepting electrons. In principle, you could identify specific compounds, such as caffeine, chlorogenic acids, or the organic acids that give coffee its brightness, by looking for their characteristic signatures on the voltammogram.
But Hendon’s team took a different approach. Instead of trying to identify individual molecules, they looked at the overall shape of the response, particularly in the region where hydrogen ions interact with the surface of a platinum electrode.
What they found was surprising.
In brewed coffee, which is naturally conductive and self-buffered to a pH of about 5, the voltammogram looks remarkably like that of acidic water. There are features corresponding to hydrogen adsorption onto the platinum surface, followed by hydrogen gas evolution at more negative voltages. On the return sweep, oxygen-related chemistry appears.
But here is where it gets interesting. When you cycle the voltage repeatedly, those hydrogen-related features shrink. The current decreases by about 34% from the first scan to the second and another 18% to the third. Something is coating the electrode surface, blocking the sites where hydrogen would normally react.
That something, the researchers discovered, includes caffeine.
Scavenging the Cup
To prove this, they did something clever. They took a platinum mesh electrode, far larger than the tiny disk used for routine measurements, and cycled it hundreds of times in brewed coffee, deliberately building up a layer of adsorbed material. Then they submerged the electrode in a water-acetonitrile solution, sonicated it to release the adsorbates, and ran the resulting liquid through a high-performance liquid chromatograph coupled with a mass spectrometer.
Caffeine showed up. About 300 micrograms of it, representing roughly 0.4% of the total caffeine in an average cup. Over the course of the experiment, each hundred-cycle scan scavenged about 0.1% of the available caffeine.
But caffeine is not the whole story. Dark roasts have less chlorogenic acid than light roasts. Those compounds break down during roasting, contributing to the bitter, smoky, “dark” flavor profile. The team used density functional theory calculations to show that both caffeine and 5-caffeoylquinic acid, a common chlorogenic acid isomer, bind stably to platinum surfaces, with slight preferences for different crystal facets. The suppression of the hydrogen signal, they argue, reflects the ensemble of organic molecules competing for the electrode surface. And that ensemble changes with roast level.
Distilling the Data
To test this hypothesis, the researchers did something any good coffee scientist would do: they roasted coffee. Starting with a Colombian green bean, they generated six progressively darker roasts, ranging from 75.8 Agtron units (light) down to 55.7 (dark). They rested the beans for seven days to allow carbon dioxide to off-gas, then brewed them using the Specialty Coffee Association’s cupping protocol.
Here is the critical step. They diluted each brew to exactly 1.00% TDS, measured by refractometer. So all six coffees had the same strength. Any difference in the voltammogram would therefore be due to composition alone, to roast level.
The difference was dramatic. The lightest roast passed about 50% more charge in the hydrogen region than the darkest roast. When they plotted total charge against TDS for each roast, they found a linear relationship, but the slope was steeper for lighter roasts.
In other words, the electrochemical method can decouple strength from roast color. Two coffees with the same TDS but different roast levels produce different electrical signatures. That is something a refractometer cannot do.
The Blind Taste Test
But the real validation came from a collaboration with Colonna Coffee, a specialty roaster in Bath, UK. Colonna had roasted four batches of the same coffee to the same target whole-bean color, about 93 Agtron units. Three of the batches were acceptable. One was rejected by their sensory quality control panel because it was too light, 98.9 Agtron, and exhibited undesirable flavors.
The roaster sent the samples to Hendon’s lab in single-blind fashion: four unlabeled samples, no indication which was rejected.
The team brewed each sample five times, in random order, and ran their voltammetry measurements in another random order. The refractometer readings showed no statistical difference between any of the four samples. The whole-bean color measurements, the very specification the roaster was trying to hit, could not distinguish the rejected batch from the acceptable ones.
But the electrochemical method could.
The current passed in the first scan clearly separated sample 1, the rejected batch, from samples 2, 3, and 4. The differences were statistically significant, with p-values as low as 0.0002. The acceptable batches all fell within the same statistical class.
The fouling rate, how quickly the current decreased from scan one to scan two, was identical across all four samples. That rate depends on concentration. But the absolute current in the first scan depends on composition. By looking at the first scan alone, the method correctly identified the out-of-spec coffee.
The roaster confirmed: sample 1 was the rejected batch.
Why This Matters for the Coffee Industry
Let me pause here and translate what this means for someone running a roastery or a café.
Right now, quality control is a patchwork. You measure bean color with a spectrophotometer. You measure brew strength with a refractometer. And then you taste. But tasting is subjective, and even the best palates fatigue. A batch that passes all the instrumental checks can still fail on the cupping table because something subtle went wrong in the roast, a slightly uneven development, a minor deviation in the temperature curve, a bean that did not behave the way the previous batch did.
The electrochemical method offers something new: a single measurement that captures both the amount of coffee in the cup and the kind of coffee that is there. It is sensitive to the ensemble chemical composition in a way that refractive index is not.
Hendon’s team envisions quality control calibration curves. A series of simple CV measurements on progressively more dilute coffee allows a roaster to rapidly construct a reference, enabling quantitative comparisons of separate batches of the same coffee roasted to the same color.
But perhaps more intriguingly, the method is sensitive to differences that even color-matched batches can show. Those four batches from Colonna had nearly identical Agtron readings. The refractometer could not tell them apart. The human tongue could, but the electrochemical method could, too, and with quantitative precision.
What the Method Cannot Do (Yet)
A responsible reporter must also note the limitations.
First, the method requires a potentiostat and a platinum electrode. While these are not exotic instruments, potentiostats are common in electrochemistry labs and are becoming smaller and more affordable, they are not yet a café countertop tool. The researchers have a financial interest in a company called Overpotential, which is working to commercialize electrochemically modified food products, suggesting that they see a path to real-world application. But we are not there yet.
Second, the method does not replace tasting. It supports it. The goal is not to build a machine that tells you whether a coffee is “good” or “bad” in some absolute sense. The goal is to build a machine that tells you whether this batch matches the chemical profile of the batch you approved last week. Consistency, not judgment.
Third, the research was conducted on a relatively narrow set of coffees, a single Colombian origin roasted to different levels, plus a validation set from a roaster in the UK. The authors acknowledge that the shape of the “plane” mapping charge to TDS and Agtron color may be coffee-specific. A robust quality control system would require calibration curves for each coffee, each roast profile, each brewing method.
And finally, the method as currently described requires the coffee to be brewed to cupping standards, a standardized protocol that includes a specific water temperature, contact time, and filtration method. Real-world brewing in a busy café is messier than that. Whether the method remains reliable across variable grind sizes, water compositions, and brewing devices is an open question.
The Deeper Insight
But there is something deeper here, something that speaks to a broader shift in how we think about coffee quality.
For decades, the specialty coffee industry has pursued a kind of analytical reductionism. We measure TDS. We measure extraction yield. We measure bean color. We measure particle size distributions. We track water chemistry to the part per million. The implicit goal is to control every variable so precisely that the sensory outcome becomes predictable.
But coffee resists that kind of control. Not because we lack precision instruments, but because the relationship between the variables and the sensory experience is nonlinear, emergent, and deeply dependent on the ensemble chemistry of the brew.
What Hendon’s team has done is to embrace that complexity rather than try to reduce it. They are not measuring individual compounds. They are measuring the collective effect of those compounds on a simple electrochemical process, hydrogen adsorption onto platinum. The current depends on how many protons are available and on how many organic molecules are competing for the electrode surface. That competition is a proxy for the overall chemical character of the brew.
In a sense, the voltammogram is doing something very similar to what your tongue does. Your taste receptors respond to patterns of molecular activation, not to individual analytes. Sweetness is not sucrose; it is the activation of a family of receptors by a range of molecules that share certain structural features. Bitterness is similarly complex. The electrochemical method captures a related kind of ensemble property.
This is not a coincidence. Both taste and electrochemistry are fundamentally about molecular interactions at surfaces.
A New Tool for an Ancient Craft
Coffee has been drunk for at least 500 years, and for most of that history, quality assessment was purely sensory. You tasted it. If you were good, really good, you could identify origin, roast level, and defects by smell and taste alone.
The modern specialty coffee movement has added instruments to the toolkit: color meters, refractometers, moisture analyzers, gas chromatographs. Each has improved consistency. Each has also revealed new dimensions of variability.
The electrochemical method proposed by Hendon’s team is the latest addition to that toolkit. It is not a revolution that renders the human palate obsolete. It is a new lens that reveals something the other lenses miss. It sees composition where the refractometer sees only concentration. It sees the difference between light and dark that a spectrophotometer, fixed on a single color target, can miss.
And in a blind test against a roaster’s own quality control panel, it got the answer right.
That is the standard that matters. Not whether the method is elegant or novel or scientifically interesting, though it is all of those things, but whether it can do work that needs doing. Whether it can help a roaster catch a bad batch before it goes out the door. Whether it can help a café reproduce a beloved brew day after day. Whether it can give the coffee industry something it has never had: a direct, quantitative, in-situ measurement of the chemical properties that actually determine flavor.
The answer, based on this study, appears to be yes.
The Bottom Line
Christopher Hendon and his colleagues have shown that cyclic voltammetry can measure both the strength and the roast level of black coffee in a single, rapid test with no sample preparation. The method is sensitive enough to distinguish between batches of coffee that have identical TDS and nearly identical bean color, batches that a refractometer cannot tell apart and that a roaster might reject only after tasting.
This is not yet a café-ready tool. But it is a proof of concept for a fundamentally different approach to coffee quality analysis: one that measures ensemble chemical properties rather than individual analytes, that embraces complexity rather than reducing it, and that aligns more closely with how human sensory perception actually works.
For an industry that has long sought a quantitative method to assess beverage qualities beyond those informed by sensory panels, this is a significant advance.
And for the rest of us, the millions of people who start each day with a cup of coffee that is sometimes transcendent and sometimes merely adequate, it is a reminder that the science of that morning ritual is still being written. The perfect cup is not yet a solved problem. But we are getting closer.
The study, “Direct electrochemical appraisal of black coffee quality using cyclic voltammetry,” appears in Nature Communications (2026, Vol. 17, Article 3618). Christopher H. Hendon and Doran L. Pennington have a financial interest in Overpotential, a company commercializing electrochemically modified food products.
