The Chemistry of Roasting: Maillard, Caramelization, and the Acids

A roasting coffee bean is one of the busiest chemical environments in your kitchen’s supply chain. Over the course of a ten-minute roast, hundreds of reactions run simultaneously and interdependently inside each seed, converting a stockpile of flavorless precursors — sugars, amino acids, organic acids, lipids — into the more than 800 volatile aromatic compounds identified in roasted coffee. The roasting process describes this journey by its visible and audible milestones; this article opens the hood and looks at the reactions themselves. There are four big ones worth knowing — Maillard, Strecker, caramelization, and pyrolysis — plus a supporting drama involving coffee’s acids that explains, among other things, why dark roast is bitter and light roast is bright.

The Maillard reaction: coffee’s flavor engine

If you learn one piece of food chemistry in your life, make it the Maillard reaction — the non-enzymatic browning that occurs when amino acids and reducing sugars collide under heat. It’s the reaction that crusts bread, sears steak, browns onions, and, in coffee, does more to create flavor than any other single process. Maillard chemistry gets going meaningfully around 140–165°C of bean temperature and continues right through first crack, and it isn’t one reaction so much as a cascade: initial sugar-amino couplings rearrange and fragment into a sprawling family tree of products, each generation more aromatic than the last.

Two outputs matter most. The first is the aromatics: furans that smell of sweet caramel, aldehydes that read as fruity or malty, diketones like 2,3-pentanedione that smell of butter, pyrazines that give roasted, nutty, earthy notes. The second is melanoidins — large, brown, nitrogen-containing polymers that give roasted coffee its color, contribute substantially to body and mouthfeel, and even carry antioxidant activity into the cup. When you notice that a coffee feels weighty on your tongue, you are, in part, tasting melanoidins.

The Maillard phase is also a control surface. Run through it quickly and the resulting cup keeps more acidity and vivid sharpness; linger in it and the cup gains rounder sweetness and heavier body. This is why two roasters with identical beans and identical end temperatures can still produce noticeably different coffees — the path matters, not just the destination.

Strecker degradation: the quiet collaborator

Riding along with Maillard is a less famous partner, Strecker degradation, in which free amino acids are dismantled — in the presence of oxidizing compounds that the Maillard cascade itself conveniently supplies — into aldehydes, ammonia, and carbon dioxide. The aldehydes are aromatic heavyweights: methylpropanal and 3-methylbutanal, for instance, are responsible for malty notes, and the broader Strecker family contributes fruity, floral, grassy, and nutty aromas. Because Maillard chemistry feeds Strecker chemistry, a roaster who extends the Maillard phase is also, knowingly or not, setting up more Strecker activity later in the roast — a nice example of how these reactions interlock rather than queue up politely.

Strecker degradation has one more claim to fame: the CO₂ it generates is a major contributor to the gas pressure that eventually fractures the bean at second crack, and to the degassing that continues for days after the roast.

Caramelization: sweetness spent, complexity gained

Caramelization is browning of a different kind — no amino acids required, just sugars and heat. Sucrose, green coffee’s dominant sugar, begins breaking down somewhere around 160–180°C (the threshold shifts with how fast the temperature is climbing), splitting into simpler sugars and then fragmenting into a huge cast of ketones, esters, aldehydes, and furans whose aromas span caramel, malt, toast, and fruit.

Here lies one of roasting’s best-known ironies: caramelization does not make coffee sweeter — it makes it taste caramelized while making it less sweet. Every sugar molecule that browns is a sugar molecule destroyed, traded for complexity and color. A little of that trade builds the toffee-and-toast depth we love in a medium roast; run the trade too long and you’ve spent the entire sweetness budget on increasingly bitter, roasty compounds. This is the chemical basis for sweetness following a bell curve across roast levels: underdeveloped roasts haven’t converted enough, well-developed mediums sit at the peak, and dark roasts have burned through their principal.

Pyrolysis: where roasting becomes burning

Past second crack, chemistry gives way to demolition. Pyrolysis — thermal decomposition of the bean’s material itself — takes over as the dominant process, charring carbohydrates and cellular structure, pushing melanoidin formation to its extreme, and generating the smoky, carbonized flavors of the darkest roasts. A managed brush with pyrolysis yields the heavy body and bittersweet punch of a French roast; an unmanaged one yields ash. At the far end, the bean is literally carbonizing — losing its oils, aromatics, and solubles to smoke — which is why extremely dark coffee tastes not stronger but emptier.

The acid story

Coffee’s acids deserve their own ledger, because roasting treats them very unequally — some are casualties, some are products, and one is a troublemaker in both directions.

Citric and malic acids are formed in the coffee cherry as it develops on the tree — they’re gifts from the farm — and heat simply spends them down. This is why light roasts, which end before the destruction gets far, taste of citrus and apple in a way dark roasts never do. Acetic acid runs the opposite direction: it’s manufactured during the roast as sucrose fragments, and in the right dose it contributes a pleasant winey vivacity (too much reads as harsh and fermented).

The most consequential entry is the last one. Chlorogenic acids — a family of polyphenols abundant in green coffee and prized for their antioxidant properties — degrade progressively with heat, and their breakdown products include quinic and caffeic acids, both bitter. So dark roasting performs a double subtraction: it destroys the bright, fruity acids while simultaneously generating bitter ones. A light roast may retain a substantial share of its original CGAs; a very dark roast loses the overwhelming majority. That single ledger line explains most of the acidity difference you taste across the roast spectrum.

Why any of this matters at the brewer

You don’t need a chemistry degree to make good coffee, but this chemistry explains things every brewer eventually notices. It explains why light and dark roasts of the same bean taste like different beverages: they contain measurably different inventories of acids, sugars, and aromatics. It explains why light roasts are harder to extract — the structural breakdown driven by these reactions hasn’t gone as far, so the bean remains dense and reluctant. It explains why over-extracted coffee turns bitter, as the slower-dissolving bitter compounds (quinic acid and friends among them) catch up with the fast-dissolving bright ones. And it explains why the roaster’s decisions are permanent: whatever balance of reactions was struck in the drum is the palette you brew from. Grind and technique can present that palette well or badly, but nothing in your kitchen can put a destroyed acid back together.

The bean, in other words, arrives at your grinder as a finished chemical argument — hundreds of reactions, one conclusion. Brewing is just the reading of it aloud.