Measuring the Water activity of Fine Coffee Raw Bean (part I)

For professional baristas, please follow the coffee workshop (Wechat official account cafe_style)

For the coffee industry, it is a well-known fact that the quality of raw beans declines over time. The demand for high quality of boutique coffee makes the industry more sensitive to the issue of raw bean aging.

One of the biggest challenges for importers is how to maintain the original quality of the top raw beans purchased during transportation. The impact of long sea and land trips on raw beans may be bad luck. When a cup of expensive raw beans with a score of 92 points arrives at its destination with only 87 points, the beans can no longer be among the top boutique coffees. Those unique and delicious cups are the real reason why we are willing to pay high prices for coffee farmers.

A benign industrial chain should be like this in theory.

Good coffee allows farmers to sell at a better price, produce more good coffee, and then once again allow us to buy high-quality raw beans at high prices. This means that the bean baker has more opportunities to present the level of coffee, and the store can also provide delicious drinks for the guests.

Sam Miller is assembled in Finca Pennsylvania

Premature aging of raw beans is a problem that can occur at any stage. Repeat selling and high pricing strategies are risky for producers because cup test scores directly affect the purchase price of high-priced beans, which is no longer at the level of the coffee futures market plus a basic premium. Both will be important considerations. At the same time, if coffee farmers put considerable effort into producing high-quality coffee, but only get the quality of commodity beans, if we assume that the price is based on cup quality, farmers will certainly not be able to make ends meet.

For importers of raw beans, especially boutique importers who do not use the category of futures coffee, if the condition of raw beans continues to decline, resulting in a batch of coffee must be based on the futures price as the benchmark price, these cumulative losses will soon become very considerable. For bakeries and coffee shops, the brand image based on quality and technology is important, and the loss caused by aging beans will evolve into "ingenious use" such as their own specialties or being deeply baked in French style. but it will also lead to thorny conflicts with guests or awkward silence.

In order to understand the aging process of raw beans, Cafe Imports, an importer of raw beans in Minneapolis, USA, began to explore the context of all this from the perspective of food science. Over the past year, we have worked with Decagon Devices, the developer of the humidity meter, to study what we consider to be a key measure: water activity (Water Activity), which may be an indicator of the aging rate of raw beans and may help prevent a decline in raw bean quality in the future.

Huila coffee dryer assembly equipment

Although industry pioneers such as Cup of Excellence and Terrior have measured the water activity of raw beans in the past, the use of this data to predict the stability of top-quality bean flavor has not been firmly proved. Previous studies on coffee and water activity have not involved the loss of flavor in terms of quality, let alone the difference in quality quality from 92 points in the cup test to 87 points in the light fragrance. 87 points can definitely be counted as boutique coffee, but the bulk standards of commercial beans or food safety levels do not care about these subtle differences, and current research is limited to maintaining minimum standards and inhibiting microbial reproduction in food.

But for us, the meaning of water activity is completely opposite. We rarely see data on the activity of coffee water within the range of microbial proliferation, and even if there are, most of them are batches that have been eliminated. So the crux of the problem is that when coffee is bought with a top boutique rating of 92 points and can only be sold with the quality of 87 points of boutique commercial beans, the situation is completely different. The direction of our research is not to verify the relationship between water activity and coffee in the field of futures commodities, but to summarize the application of data to a more specific area; the principle of water activity remains the same, but is no longer limited to the application of commercial food safety standards.

Water activity (aw) is the ratio of equilibrium vapor pressure (P) to saturated vapor pressure (p0) of pure water: aw = p/p0. Simply put, it measures the energy state of water in a system. Water activity is different from water content, which, as the name suggests, MC measures the proportion of water contained in a system. These two concepts can be explained by the total amount of water molecules, bound water molecules, and free water molecules. Moisture content takes into account the sum of all the moisture in raw beans. But the properties of water are not necessarily the same in a substance. The water content of any substance at the cellular level is more or less limited by its properties. Water activity measures the degree of this dependence, not the ratio of water molecules to free water molecules in water content. Water activity is not a numerical value of content, but a measurement of energy, and whether water is bound to matter in a system. Or it exists in the form of flow.

We used 4TE   Duo, a water activity measuring instrument developed by Decagon   Devices's AquaLab department. There are two ways to explain the operation of this instrument:

4TE Duo Dew Point humidity tester for measuring water activity

My point: first of all, put the sample in the sample cup. Next, put the sample cup into the instrument, and then place the joystick to "read". The third step is simple: start waiting. One of the disadvantages of the machine's sensitivity is that the whole step will take at least two minutes, but it can also take as long as 20 minutes (while the moisture meter only takes a few seconds to measure). Usually in about five minutes, the instrument will make a sound after reading the stable data, and we will record the results.

Drying equipment of Finca Pennsylvania

Dekegang's official statement: the best way to measure water activity is recommended to be measured in the proportion of p/p0. The saturated vapor pressure (p0) of pure water depends on the sample temperature. If we know the temperature of the sample, we can get the value of p0. The moisture vapor pressure of the sample can be measured from the space vapor pressure at the top of the seal in the instrument. But back to the principle of physics, the most accurate way to measure saturated vapor pressure is to measure the dew point of the air. The AquaLab dew point meter uses dew point temperature to measure vapor pressure, so it provides a direct and basic measurement of water activity.

In addition, Dirk just described it this way: "the far-infrared beam focused on the micro-mirror can judge the precise dew point temperature of the sample." This dew point temperature is then used to determine water activity. "

If we can judge and measure the strong correlation between water activity and the rapid aging of high-quality coffee beans, we may be able to gradually lock in what conditions make the situation of raw beans begin to decline. It also means better quality of raw beans, higher prices, and better coffee brewing. This time it is our new idea to apply the measurement of water activity to the analysis of high-quality coffee. The first goal of this study is to uncover the reasons for the decline in the cup test score of a bean and how this change limits the price of coffee; the second goal is to explore whether the state of raw coffee beans can be improved by analyzing water activity before transportation. The question here is very simple: if we find that a sample cup before shipment has 92 points and will fall to 87 points after delivery, then how many samples with 87 points before shipment have fallen from 92 points? Then the question is: can we prevent this situation?

Our research design must start with a simple blueprint. Although there are data on water activity and boutique coffee, few studies have compared the two. First of all, we assume that water activity provides a reliable prediction for the stability of high-quality coffee beans. On this basis, our second assumption is that in the drying and pre-cabinet steps of the raw bean processing process, the water activity provides a measurable and most practical system to help any batch of beans stay at their best. This is very different from simple acceptance or rejection in the futures market, and it has nothing to do with whether the reproduction of microbial mold will break out in the raw bean warehouse. What we want to know is, whether the fruit-like acidity and sweetness, delicate floral aroma and caramel finish of the beans originally contained in the pre-shipment samples will still be preserved after arrival.

Aw: traditional usage and New Application

The concept of water activity is not new. In 1953, William James Scott first confirmed that the reproduction of microorganisms in food mainly depends on water activity rather than water content. What does this have to do with the water activity of coffee? Water activity is the same for coffee as for all other substances. The 24-page table lists common foods and corresponding water activity, as well as microbes that multiply at all levels. Dried fruits have much higher moisture content than oatmeal, but oatmeal has higher water activity than dried fruit. Assuming that the aw of dried fruit is raised to 0.65, some drought-resistant molds can proliferate. Even if the water content is high, this will not happen unless the water activity increases. The condition of microbial proliferation was finally dominated by water activity, which was a breakthrough in 1953. Today, this is still the scientific standard.

For coffee, measuring water activity or moisture content depends entirely on what you want to know.

If you want to judge the proportion of the purchase price of raw beans is spent on the weight of water, moisture content is a relatively ideal data. But if you need to study the possibility of preservation or loss of raw bean water, or the possibility of microbial proliferation, water activity can provide better information. In the latter case, the water activity is the actual measurement. For the assessment of water transfer and microbial reproduction, the indicators given by moisture content are really insufficient. In other words, water content can be used as a defect prediction of water activity. If the moisture content is used to predict the dynamics affected by water activity (including the aging of flavor), the errors of these predictions will be doubled. But even so, the indicator of water activity is not omnipotent, and it is certainly not like that of science fiction films. For example, water content, water activity is also measured at a certain point. We are studying how to use this value to warn the fluctuation of the condition of raw beans. whether it is estimated or referred afterwards, we cannot know from the fixed point whether the water activity has once reached 0.6889 during transportation, nor can we know how long it lasted even if it was soaring.

William   James   Scott found that water activity is the dominant factor affecting microbial reproduction, because one of the basic functions of water activity is the transfer direction of water in the dominant substance. The movement of water will eventually reach dynamic balance. Dirk Gang said that the direction of water transfer is to move from areas with high water activity to areas with low water activity, and eventually tend to energy balance. In the state of low water activity, that is, in the absence of free water molecules, water cannot pass through the cell membrane of microorganisms, making the cells latent. In the case of high water activity, compared with more free water molecules, water not only can penetrate the cell membrane, but also has a tendency to move.

We asked Wendy   Ortman, a product experimental technology engineer at Dirk Gang, to learn more about the mechanism of water transfer and water activity. She explained that the addition of soluble substances would reduce water activity. "the molecular effect makes the solute hinder the kinetic energy state of water, but this effect also affects the interaction between water molecules and solutes. Sugar and salt can reduce water activity very effectively because these substances bind to water molecules. "

Nowadays, water activity is widely used to regulate food safety. How to inhibit the water activity of various microorganisms has also been related to improve the research, this set of control system is to formulate and maintain food safety standards. The 24-page table, which has just been authorized by Dirk, lists common microbes and related water activity inhibition criteria.

Where will coffee be listed on this form? At this point, it must be pointed out that the food listed on the right side of this table actually corresponds to a range of water activity. The activity of raw bean water measured by us can be as high as 0.7549, while the lowest degree is 0.3149. Most of the coffee is in the middle range (0.5m / m / m / 0.6), showing a relatively normal distribution so far. In addition to numerical differences in water activity itself, these data also change over time. Our study will do static analysis based on single data recorded by individual raw beans, as well as analyze how these data change over time.

Water activity is also used in food design to enhance the preservation of food taste and taste. Two good examples are raisin cereal and cupcakes, both of which are made up of two ingredients with different taste characteristics. Water activity makes it difficult for the properties of these ingredients to last.

At the beginning of the raisin cereal, the raisins are chewy and slightly moist, and the cereal is crisp. But soon the raisins became dry and hard, and the cereal became tasteless. The high water activity of raisins transfers its original moisture to cereal with low water activity, creating a completely unwanted taste.

The frosting on the cupcake contains more water than the cake itself, but the sugar in the frosting binds to these water molecules much stronger than the flour in the cake can attach to the moisture. With the passage of time, the cake will soon dry out. Although the cake has low water content and relatively high water activity, it transfers water to frosting in the process of returning to equilibrium. In both cases, the process of dynamic equilibrium continues until the water activity (rather than moisture content) of the two components in the food reaches the same level.

Installation technician Sam Miller working scene in Pennsylvania Manor

The stability of food safety and quality preservation is a typical application of water activity. Our research on raw coffee beans has also analyzed these two applications. According to the same principle, using this data, the next reasonable step is to try to predict the stability of cup quality and cup quality.

But what about raw beans? How does water activity work when there is only a single substance? Even unfrosted raw beans still have two primary factors: the environment and the water activity of all raw beans. As far as the impact of the environment is concerned, air is considered. As far as coffee is concerned, the difference in a single batch is considered. These two factors can cause the overall water activity to be out of balance, in other words, in a state of activity.

We believe that the aw value can be used to predict the quality, stability and unavoidable or early aging of raw beans. Water activity is completely different from the measurement of water content, it measures the value of energy. Suppose the moisture content of a bean increases from 11.5% to 13.5%, and then falls back to 11.5%. What is lacking in these data is that the process of raw beans cannot be fully explained. Water activity aw values also fluctuate with time, but perhaps because these fluctuations are energy dynamic processes, the current water activity data do not seem to increase or decrease sharply. If we can use aw as the basis for measuring the drying process and quality preservation cycle of raw beans, there may be stronger evidence for trajectory analysis of raw bean quality.

The data of water activity were used in this study based on the known effects of water transfer, relative humidity and enzyme activity on the aging of raw beans. Raw coffee beans still have the respiratory characteristics of the plant itself. Jean   Nicolas   Wintgens states in the reference book "Coffee:   Growing,Processing, Sustainable   Production" that respiration uses oxygen in the air to produce an exothermic reaction between carbon dioxide and water by using the organic matter (starch, carbohydrates, lipids and proteins) in beans as energy and decomposing enzymes to produce carbon dioxide. These organic substances are the source of the flavor and aroma of coffee. Respiration consumes these organic substances as energy for oxidation, which is the primary factor in the aging or deterioration of raw beans. Respiration circulates constantly, and once catalyzed, it is easier to continue to self-execute. This means that the respiration of the cell produces a series of more respiration until the energy is exhausted.

The key catalysts for respiration are the relative humidity, moisture content, temperature and air composition of the environment. Water activity controls the interactive transfer between environmental relative humidity and moisture content of raw beans. At the same time, the enzyme activity and the rate of oil oxidation decreased with the decrease of water activity. Even if water activity is not the only reason for the deterioration of coffee quality, it is at least directly driven or seriously affects the decline in the preservation of raw beans.

Further investigation: resources used and methods adopted

First of all, I would like to make it clear that we are not a specialized research institution. Although we hope to lay the foundation for future experiments, we are still in observational research at this stage. Our data come from a relatively wide range of uncontrollable sources; the measured raw beans may be dried on the balcony, or under plastic sheds or viaducts, or mechanically dried; many varieties grow in very different environments and are processed in a unique way; beans are also exported by sea or land after being stored at different times.

The most amazing thing about raw beans is the unique flavor of each bean and the unique characteristics created by different producing areas and cultivation methods. For our research, these unique features have all become a large number of variables.

After receiving the raw bean sample, we will arrange the cup test within a few days in order. The day before the cup test, the sample will be unsealed, measured and baked. At this stage, the degree of water activity was recorded for the first time. At the same time, the newly purchased raw beans arrive and enter the warehouse; at this time, another sample will be taken to measure the status of the raw beans and the cup test. We record moisture content, water activity, cup test scores and all related notes, and record any grading and defects of raw beans when preparing the baking sample.

The data we collect can be analyzed at different levels from simple to complex. The simplest question is: what is the relationship between water activity aw and cup score? In other words, if all the aw data are analyzed with the corresponding cup score, will the cup score fall in a certain range of water activity? Does that have anything to do with moisture content? The scatter chart above lists the data of our laboratory for nearly a year, in which the cup score is shown on the Y axis, and the corresponding degree of water activity is on the X axis.

The above chart is the foothold from which we can construct a more complete hypothesis and divide the distribution blocks of the samples more clearly.

The cup score of coffee is affected by many factors. If we categorize the data accordingly, we can assume different situations and ask more questions, such as: how many coffee cups with high water content and high test scores (in the upper right corner of the chart) maintain high scores? And the same question, what if it is applied between moisture content and cup test score?

0.6100 and 0.6050 of the X axis on the scatter map are marked in red. This is actually about 12% moisture content, but the relationship is not proportional after all, so 0.6050 of the water activity can have 12.8% moisture content, and in another bean it may only have 11.8% water content. Of course, the water activity of raw beans with 11.8% water content may be 0.5600 or 0.6200. Therefore, there is not a proportional relationship between water activity and water content. From the distribution shown in the chart, it seems that there is no regularity between water activity and cup score. However, after further analysis and differentiation of these samples, we can summarize the internal relationship between cup test performance and these data.

In evaluating the relevance of the data, we notice the marginal data on the chart, that is, the interval value between the normal distribution and the outlier. When the water activity is more than 0.6500, the moisture content is often more than 12%, usually that raw bean has not been considered. But back in the area of the red line on the chart, the situation becomes interesting.

One of the ways we evaluate these data is to consider the accuracy of cup scores and raw bean grading. In this way, we can have a unified standard for scoring and a larger sampling group. At present, on the whole, the prediction of cup quality by water activity corresponding to the grade of raw beans is relatively accurate than that by moisture content. With the water activity mediated by the scatter map between the two green boundaries, coffee is more likely to measure the score expected by its classification, and this correlation is better than considering the appropriate moisture content corresponding to it. However, when the water activity is between 0.6050 and 0.6100, there is a significant difference between the sample cup test scores, and the water activity and water content do not seem to be able to reach a consensus. In this interval, according to the accuracy of the classification, the correlation still tends to the degree of water activity. When the water activity is less than 0.6100 and the sample cup test score of the sample whose moisture content is higher than 12% score 74% meets the grading standard, but in the same corresponding group, the water content is only 51% accurate (aw > 0.61, MC < 12%).