What Happens Inside the Roaster?

Phenomenon

Pour a handful of green coffee into a roaster.

For the first few minutes the smell is grassy, then hay-like, and the beans slowly turn from grey-green to yellow. A while later the smell becomes bread and nuts, and the color keeps deepening. Then the drum fills with a crackling sound, like popcorn.

Weigh the beans when they come out: they are lighter. Look at them: they are visibly plumper.

Where did the weight go? Why did the beans get bigger? And what was that crackling?

Variables

Roasting can be made very complicated, but this chapter looks only at the most basic dials:

This chapter is about what happens inside the bean, not how to operate a roaster. A real roast profile involves far more than this, and we do not pretend to simulate one here.

Experiment

Thought experiment: weigh out two equal batches of green coffee and roast them in a transparent drum. Take one out at a light color; take the other all the way to dark.

Write down your prediction before reading on

When they come out, which batch weighs more? Which takes up more space? In what order will the colors change? If you break a bean in half, will the inside be the same color as the outside?

Your prediction

Save prediction

Stored only in this browser on this device. Never uploaded.

A second experiment: roast two more batches and bring them to exactly the same final color — one with high heat and a short time, the other with gentle heat and a long time.

Write down your prediction before reading on

Will these two batches of the same color taste the same?

Your prediction

Save prediction

Stored only in this browser on this device. Never uploaded.

Observation

In the first experiment, both batches got lighter and bigger, and the dark one changed more. During roasting, beans lose water and part of their dry matter while their volume expands, their density falls and their internal pores multiply; at a dark roast, volume can grow by nearly 80 percent and about a fifth of the dry matter is lost Confidence: [Strong Evidence].

The color moves roughly along green → yellow → light brown → dark brown; roasters commonly use that color sequence to describe how a roast is progressing Confidence: [Common Practice]. Break a bean open and the inside is often a little lighter than the outside, and the faster the roast, the bigger the difference tends to be Confidence: [Rule of Thumb]. The reason is simple: heat travels gradually from the bean's surface to its center Confidence: [Established Mechanism].

The second experiment is more interesting: the two batches of the same color do not taste the same. One study roasted beans to the same degree of roast using six different combinations of time and temperature and got clearly different aroma profiles Confidence: [Strong Evidence]; another found that at the same degree of roast, beans roasted at a higher temperature had more volume and larger pores in their cell walls Confidence: [Strong Evidence].

Model

Think of roasting as a timeline driven by heat. The names of the stages below are there to make it easier to talk about; inside the bean they overlap, with no clean boundaries.

1. Drying

Green coffee still holds a fair amount of water. Early in the roast, much of the heat goes into warming the bean and turning water into steam; a large part of the water has to evaporate before the bean's temperature can keep climbing Confidence: [Established Mechanism]. The grassy smell fades and the beans turn yellow.

But drying is not a separate warm-up stage. Research found that most aroma compounds form precisely while the bean's moisture is still dropping Confidence: [Strong Evidence]. So "drying" and the "browning" below overlap in time.

2. Browning: the Maillard reaction, and caramelization

"Browning", put plainly, is the bean turning brown and starting to smell like baking bread.

The main chemistry behind it is called the Maillard reaction: when heated, sugars and amino acids (the building blocks of proteins) go through a chain of reactions that produce brown substances and a large number of aroma compounds Confidence: [Established Mechanism]. The crust of bread and the browned surface of a seared steak come from the same family of reactions.

Alongside it there is caramelization — sugar itself breaking down at high temperature. In coffee, the Maillard reaction, caramelization and, at higher temperatures, thermal breakdown all act together Confidence: [Strong Evidence], and inside the bean they are hard to separate cleanly.

The raw materials for these reactions are the sugars, proteins and acids stored in the seed (Chapter 2). As raw materials are used up, some substances keep declining: caffeine stays largely stable through roasting, while chlorogenic acids keep breaking down as the roast goes on — the darker the roast, the less remains Confidence: [Strong Evidence].

3. Gas formation

The Maillard reaction, caramelization and thermal breakdown also produce gas — mostly carbon dioxide, plus water vapor. When the roast ends, part of the gas is trapped in the bean's pores and is released slowly over the following days to weeks Confidence: [Strong Evidence].

That is why fresh beans "degas" in the bag, and why the coffee bed puffs up during a pour-over bloom (Chapters 17 and 27).

4. First crack

Steam and gas build up inside the bean while the bean's structure is also changing at high temperature. At some point the bean gives a sharp "pop" and starts to swell visibly — that is first crack.

With first crack, it helps to separate what is settled from what is not:

• Roasters treat first crack as one of the most important landmarks of a roast, using it to judge progress and plan the time that follows Confidence: [Common Practice];
• That beans expand during roasting and develop a large number of internal pores has been observed repeatedly with microscopes and measurements Confidence: [Strong Evidence];
• But exactly what triggers first crack — whether the pressure is mainly steam or carbon dioxide, and how the cell walls soften and break at high temperature — is not yet settled Confidence: [Open Question].

5. Development

The stretch of heating after first crack is called development. Browning continues, the color deepens, and the bean becomes more brittle and more porous.

The longer the development, the more flavor comes from the roast itself (roasty, bitter) and the less from the bean's original fruit acidity and floral notes Confidence: [Rule of Thumb]. Further on, a finer, quieter "second crack" appears, and oil starts to show on the bean surface Confidence: [Common Practice]. Light, medium and dark roasts are simply stopping at different points on this timeline (Chapter 15).

The endpoint is not the path

Put the two experiments together: roast degree (usually measured by color) is only a reading of where the roast ended. The same color can be reached by very different timelines.

It is a bit like the drawdown time in Chapter 27: a reading, not the whole story. Two roasts of the same color can differ in aroma profile and in the bean's internal structure.

What sets the course of this timeline is, first of all, heat input: the faster heat is delivered, the faster the bean warms up, and the earlier drying, browning and first crack arrive Confidence: [Established Mechanism]. Heat input does not directly decide "good or bad"; it decides when each stage arrives and how long each one lasts.

Roasting also changes extraction

Roasting changes not only flavor but also how easily the bean extracts. Darker beans are more porous and more brittle, and they usually extract more easily, which is why many people use a slightly coarser grind or slightly cooler water for dark roasts Confidence: [Rule of Thumb]. This connects to Phase 4: with the same recipe, a different roast level can shift the extraction.

Transfer

You cannot see inside the roaster at home, but you can read from the beans and the cup where on the timeline they stopped:

The adjustments in the table are starting points, not conclusions Confidence: [Rule of Thumb]. The rule still holds: change one variable at a time.

Misconceptions

"Dark roast has more caffeine and is 'stronger'." Roasting does not create caffeine; it stays largely stable through roasting Confidence: [Strong Evidence], and by weight light and dark roasts differ little. Dark beans are less dense, so if you measure beans with a scoop, a scoop of dark roast actually weighs less and holds less coffee. As for "strong": strength is set mainly by the brew ratio (Chapter 22). Dark roast tastes more intense because of roasty and bitter flavors — not caffeine, and not strength. One survey found that the caffeine in a single café espresso can vary several-fold from one shop to another Confidence: [Strong Evidence]: how much coffee is used and how it is brewed matter far more than roast level.

"Roasting just turns beans brown." Color is only the most visible part. At the same time, water evaporates, a large number of aroma compounds form, gas is produced, and the bean's internal structure is rebuilt.

"Same color, same roast." The same color can be reached with different times and temperatures, and the aroma and structure can differ.

"More heat, faster, is always a better roast." Heat input changes the pace of the timeline, not the quality. With too much heat, too fast, the outside of the bean can be dark while the inside has not caught up; with too little heat, too slowly, the flavor can turn flat Confidence: [Rule of Thumb]. The right pace depends on the green coffee and the flavor you want; there is no "more is better" direction.

Real Lab

There is no registered lab card for this chapter, and you do not need to roast anything yourself. Get one bag of light roast and one of dark roast (the closer they are to the same coffee, the better), a kitchen scale and a measuring spoon.

1. Look. Compare color and surface: which is dry, which is oily?
2. Break. Break or cut one bean from each and compare the inside with the outside.
3. Weigh. Take one level spoonful of each with the same spoon and weigh them. A scale with 0.1 g precision is best; if you do not have one, weigh several spoonfuls together.
4. Smell. Smell the whole beans first, then grind a little of each and smell again. Write down the words that come to mind.

Before you start, write down your predictions: which spoonful will weigh more? Which bag will show a bigger difference between inside and outside?

If you can buy green coffee, add it too: it is harder and heavier, and smells of grass and hay rather than "coffee". Some people roast at home in a frying pan or a hot-air popcorn popper, but it is very smoky, throws off flakes of chaff, carries a fire risk and needs good ventilation — it is not a required part of this chapter.

Check Your Model

Predict. You don't use a scale; you scoop two spoonfuls of beans per cup. Now you switch from a light roast to a dark roast and keep the same number of scoops and the same water. Before looking at any result, write down: how will the strength of the cup change?

Diagnose. A friend switches to a very dark roast, finds it too bitter, and grinds coarser. It is still bitter, and now also weak and thin. What happened, and what could they change?

Explain. In your own words, explain why two batches roasted to the same color can taste different.