Aillio Bullet R1 / R2 咖啡烘焙中的美拉德階段：泛黃、加熱和風味

On the Aillio Bullet R1 / R2, I do not treat the Maillard phase as a fixed time target. What matters more is what yellowing temperature tells us about surface moisture, how heat continues moving through the bean’s layers, and whether RoR remains in a controlled decline as power and airflow are adjusted.

This article explains how I interpret yellowing, moisture, heat transfer, and the Maillard reaction together—and why simply extending the Maillard phase does not guarantee more sweetness.

In Aillio’s RoasTime software, once we press the Yellowing button, the software automatically labels the period between Yellowing and First Crack as Maillard.

The Maillard reaction is not a single chemical reaction, but a complex series of reactions that begins when reducing sugars react with amino acids or other amino-containing compounds under heat. The initial products continue to rearrange, break down, and polymerize, while also interacting with pathways such as Strecker degradation, eventually producing a large number of compounds with different aromas and properties.

Among these compounds, pyrazines are often associated with roasted and nutty aromas, furans can contribute sweet, caramel-like, or bread-like notes, and Strecker aldehydes may produce malty, honey-like, nutty, or cocoa-like characteristics. In the later stages of the reaction, brown high-molecular-weight compounds known as melanoidins are also formed. These contribute to bean color and may also influence mouthfeel and other sensory characteristics. However, coffee flavor is not determined by any single compound. It depends on the types, proportions, concentrations, and odor thresholds of many different compounds.

The Maillard reaction in coffee does not suddenly begin at the exact moment of yellowing, nor does it occur only during the so-called Maillard phase. Temperature, moisture, pH, the availability of reactants, and the extent to which heat has reached different layers of the bean all influence the reaction rate and the compounds that are formed. Therefore, simply extending the time between yellowing and first crack does not necessarily increase sweetness or produce a more desirable flavor.

Have you noticed that the yellowing temperature is different for every coffee? Even with the same coffee, the yellowing temperature can sometimes vary from one roast to another.

As mentioned in “The Relationship Between Moisture and Roast Color,” moisture does not have a very obvious relationship with browning itself. You have probably tasted coffee with vegetal bitterness from excessive moisture that still appeared very dark after roasting.

Fortunately, the IBTS can give us some information about the moisture condition at the bean surface. In my experience, a higher IBTS reading at yellowing usually indicates less moisture at the surface. However, this only reflects the condition of the bean surface. It does not mean that:

1. The entire coffee bean has reached the same temperature.
2. The bean core has completed drying.
3. The Maillard reaction suddenly begins at that exact moment.

Compared with washed coffees, the natural-processed coffees I roast usually reach yellowing at a higher IBTS temperature. In addition to differences in moisture, I believe this may also be related to sugar content. During drying and post-harvest processing, some sugars in natural-processed coffees may be consumed by microorganisms. As a result, reaching a similar level of caramelization or browning may require a higher temperature.

For me, yellowing temperature can also help predict first crack temperature. A lower yellowing temperature usually corresponds to a lower first crack temperature. A yellowing temperature that is too high or too low is not ideal, because the Maillard reaction needs some moisture, but too much moisture can also interfere with chemical reactions.

For the natural-processed coffees I roast, yellowing often occurs around IBTS 175–180°C, while washed coffees usually yellow around IBTS 168–173°C.

My IBTS readings are generally about 6°C higher than those reported by many other users, but this does not affect how we understand the roast. Mine is a first-generation IBTS. Although I have replaced the sensor before, its readings still seem somewhat different from those of the latest IBTS. I have also seen other users in the Aillio Bullet Users Group report a similar experience.

When comparing different batches of the same coffee, I use the timing of yellowing. When comparing different coffees, I pay more attention to the yellowing temperature.

When roasting a coffee for the first time, I use visual cues from the green beans to estimate roughly where first crack is likely to occur. This is why I pay particular attention to the yellowing temperature. In my experience, the yellowing temperature can provide some indication of the likely first crack temperature, which then influences how I plan my later power reductions and airflow increases. It can even affect how I configure future roasts of that coffee.

However, visual cues from green coffee are not always accurate. The yellowing temperature, on the other hand, is usually a much more reliable signal. That is where I see the real value of yellowing temperature.

Beginners often find it difficult to judge the color of the yellowing point. They may also treat the timing and temperature of yellowing as strict rules or benchmarks that directly determine flavor. It is easy to become overly influenced by science and numbers and assume that a roast must reach yellowing at a certain time or temperature in order to be “correct.”

Even so, roasters can still use yellowing time as a practical reference point. However, this reference should not be used to compare different coffees and different roasting machines all at once. More constructive comparisons would include:

1. The same coffee roasted under different weather conditions
2. The same coffee roasted on different roasting machines
3. Different coffees roasted on the same machine

These comparisons become much more meaningful when the reference points are evaluated together with the flavors in the cup. Remember that yellowing time and yellowing temperature alone cannot prove what state the inside of the coffee bean has reached.

In Sweet Maria’s article “Using Sight to Determine Degree of Roast,” the image labeled “4). Yellow-tan stage 6:30 – 345F (173.9°C)” shows a pale yellow color with some small whitish spots. That is close to the color at which I would press the Yellowing button.

By the time the coffee reaches yellowing, the aroma of the roast usually shifts from green-bean, grassy, or damp smells toward drier grain-like or bread-like aromas. In my understanding, the coffee bean is also closer to a rubbery state at this point. On my Bullet, this usually occurs around IBTS 170–180°C. The beans release a significant amount of water vapor around this range and then gradually become harder afterward. The section above, “Glass Transition in Coffee Roasting: How Temperature and Moisture Change Bean Structure,” explains the glassy and rubbery states of coffee beans in more detail.

In my experience, coffee beans tend to release larger amounts of water vapor around two IBTS temperature ranges during roasting: 130–140°C and 170–180°C. These are also points where roasters should pay closer attention to whether the RoR suddenly drops sharply. A sudden RoR crash is something roasters generally want to avoid. For more about the consequences of a crash, see “Flick, Crash, and Stalled RoR in Coffee Roasting.”

What we can do during the Maillard phase is, again, to follow the natural progression of the roast. In the section above, “The Blind Spot in Roast Science When It Comes to Time Control,” I explained what I mean by this concept.

In “Why RoR Should Gradually Decline,” I wrote:

“If the RoR keeps declining at a steady slope during roasting, it means moisture is continuously being released.”

This is a way of interpreting the roast that I have developed through long-term roasting experience. After entering the Maillard phase, the outer layers of the coffee bean gradually become harder as they move away from a more rubbery state. As the outer layers harden, the amount of water vapor being released gradually decreases, and the natural downward trend on the RoR caused by continuous moisture loss also becomes weaker. Therefore, if the roaster does nothing, the RoR will not automatically continue declining at the slope we want.

So, we need to take action.

To keep the RoR gradually declining on the Aillio Bullet R1 / R2, the two most direct adjustments available to the roaster are decreasing power and increasing airflow. (Please try to avoid changing the drum speed as much as possible.) But how should these two adjustments be combined? And what is the difference between controlling the roast with power and controlling it with airflow?

Many roasters keep airflow low and rely mainly on reducing power to bring the RoR down. Others maintain high power throughout the roast and gradually increase airflow to the maximum in order to lower the RoR. This high-power, high-airflow approach is often associated with attempts to imitate a Nordic roast style.

Which approach is better?

Neither is ideal on its own. Power reduction and airflow increase need to be used together.

Reducing power and increasing airflow can help bring the RoR down, but the timing and magnitude of these adjustments are highly dynamic. They depend on the coffee, batch size, heat source, airflow, and roast target. This is part of each roaster’s practical know-how.

The way decreasing power and increasing airflow are combined has a direct relationship with the final flavor. I will explain this in detail in the Member’s Section.

Many roasters spend a great deal of time discussing the relationship between different classes of molecules and flavor. I agree that there is value in discussing these topics to a reasonable extent. But if we go too far, the discussion inevitably exceeds what we can realistically understand and control in practice.

Roasted coffee contains more than 1,000 chemical compounds. What we can actually control is much more limited. Broadly speaking, I think we can influence three things:

1. How much terroir characteristics and moisture content remain available to participate in chemical reactions
2. How deeply thermal energy has reached into the different layers
3. The intensity of the thermal energy

The third point, heat intensity, affects how strongly these compounds are transformed by heat. Too much or too little heat can both lead to roast defects. In “Roast Defects on the Aillio Bullet R1 / R2,” I explain Excessive Heat and Insufficient Heat in more detail.

Cooling Matters After Drop

Coffee beans continue to retain heat after leaving the drum. Faster cooling helps preserve the roast endpoint you intended.

View The Coffee Cooling Tray