To further test the impact of Aw on roasting coffee we set up a small trial in which we altered the Aw of a single coffee in various ways.
Methods and Materials
In this trial we used four conditions: Control, Bath, Heat, and Dry.
The Control condition was just the coffee itself.
For the Bath condition we put coffee samples in trays into loosely closed containers that had water in the bottom.
The coffee samples were not allowed to contact the water.
For the Heat condition we used the same translucent plastic containers to house samples in the same trays and set them in a sunny window.
For the Dry condition we used the same setup as for the others, but this time used a layer of a desiccant (diatomaceous earth) in the bottom of the container.
Coffee samples were left in their respective conditions for one to two weeks.
Samples were then tested in six ways: moisture content, density, water activity, "constant roast profiling, " whole bean Agtron, and ground Agtron.
Moisture, water activity, and Agtron have already been discussed.
For density measure we use the same Sinar instrument as for moisture, along with the Sinar volumetric cylinder.
"Constant roast profiling" means that we set the inputs on our S roasters to the same level and ran the various samples with no changes.
Rather than roast the samples on profiles, we tracked the time-temperature profiles that resulted for each sample with this constant heat application.
With this we sought to observe how the manipulated coffees would respond to identical roast inputs.
Would the higher Aw manipulation become darker?
Would it progress through Maillard more quickly?
We began by testing 24 "base" samples of our coffee to establish a baseline for each test.
Each condition was then run with seven samples. We also created seven additional control samples, this time stacking them in the same trays and containers as the others, but away from light and heat and without any manipulation.
The next step in this trial was the roasting of the coffee samples.
Coffees were left in the S with no changes made for 9 minutes.
Results:
The Baseline samples matched the Control group in our tests and we can look at the standard deviation of the Base/Control samples to help get a picture of how much each of the other treatments impacted the coffee.
The Bath condition was by far the most impactful on the test coffee's moisture and water activity levels.
We got slight decreases in both with the Dry condition.
The Heat condition was possibly restricted by being run during a cloudy period.
The Heat condition could be improved by the use of a heating pad or sunnier period of time.
The Dry condition could be improved by the use of a stronger desiccant, or possibly by the use of a greater quantity of diatomaceous earth.
For the Bath condition we found that the rate of rise was lower than it was for the other samples (which all performed similarly, though the Heat condition did turn around more quickly and rise
just slightly faster through the Maillard phase of the roasts) for the first 6 minutes of the roast.
For the 6th, 7th, and 8th minutes of roasting, the rate of rise of all samples was similar.
We found that not only was the rate of rise throughout the Bath roast slower than in the other conditions, but the Bath samples specifically were about a minute slower through the nominal Maillard phase (from around 135°-175°C) than the other samples.
In this case, average Bath condition Aw was 0.6636 (0.6850- 0.64.
This should have been right in the prime zone for either increasing the rate or the degree of nonenzymatic browning.
It could be that the corresponding increase in moisture slowed the browning progression, and that the Aw was coincidence.
In order to test for this we would need a way to increase Aw independently of moisture.
Whole Bean and Ground Agtron
The total browning of the Bath samples was considerably less than in the other conditions.
The Bath samples were less browned than the others, both when measured whole bean and when measured ground.
The Bath samples also had the greatest
difference between whole bean and ground measurements.
The rate of rise and temperature charts show the higher moisture and Aw samples maintaining a cooler roasting environment, despite being subjected to the same inputs.
This could be due to an evaporative effect.
The lighter roast output for both Bath Agtron scores, along with the significantly larger gap between those scores for the Bath samples points to a change in the larger roasting dynamic of these samples.
The Bath samples roasted differently.
The most likely explanation is that there was enough moisture present to impact the roasting environment, as seen above. However, we cannot rule out the possibility of the higher Aw having a more basic impact on the coffee itself.
Reaction rates for things like oxidation and enzymatic activity increase at higher Aw values in storage.
These could directly act on roast browning compounds in coffee while in storage.
Further and more complex testing would be required to explore this hypothesis.
Given the same roaster input conditions, it appears that higher Aw and moisture coffees may brown more slowly and to a lesser degree.
Given the same roaster profile conditions, it is difficult to see a pattern.
What pattern does appear seems to suggest that coffees beginning with higher Aw and moisture may resist browning.
Looking at both, it may be that these coffees tend to require more energy to achieve the same degree of browning as lower Aw and moisture coffees.
출처 : CAFE IMPORTS
