Food Chemistry




There are four parts to the "Water" section

Water activity and moisture sorption


The key points to get in this section are

Water activity (aw)

aw of frozen foods

Moisture sorption

Typical shapes

Zones of isotherms

The BET monolayer

Food spoilage

Multicomponent foods

Most of the material is covered in Fennema section 2.9 and 2.10

Water activity (aw)

Drying is a key aspect of food preservation, and almost traditional processes (e.g., salting, freezing) in some way alter the properties of water in foods.   We are looking for a unified way of explaining this that we can use to predict how well our foods are preserved.  Water content is clearly not a directly useful parameter (frozen and dried beef are very shelf stable but have different water contents) so we need to think of something else.   there are two main approaches water activity/moisture sorption (this section) and molecular mobility/glass transition (next).

Water activity is a measure of how efficiently the water present can take part in a chemical (physical) reaction.  If half the water was so tightly bound to a protein molecule that it could not take part in a hydrolysis reaction the overall water activity would be reduced.  It turns out that


where p and po are the partial pressures of water above the food and a pure solution under identical conditions respectively.   The tightly bound water has no tendency to escape from a food as a vapor and therefore exerts no partial pressure and has an effective water activity of zero.    The amount of water binding is therefore indicated by the water activity.   Water activity is clearly a function of composition but is also a function of temperature.  The approximate water activities of some common foods are given below:

1-0.95   Fresh fruit, meat, milk
0.95-9   Cheese
0.9-0.85   Margarine,
0.85-0.8   Salted meats
0.8-0.75   Jam
0.75-0.65  Nuts
0.65-0.60  Honey
0.5   Pasta
0.3   Cookies
0.2   Dried veg., crackers

The water activity of a food in equilibrium with its surroundings is also defined as the equilibrium relative humidity of the surrounding atmosphere (a fraction).


aw of frozen foods

One immediate problem with water activity is that it is not useful for frozen foods.  The water activity of a frozen food is equivalent to the partial pressure over the food divided by the partial pressure of supercooled water at that temperature (measurable to about -15oC).  The partial pressure for the frozen food is equivalent to the partial pressure of ice at that temperature (water in foods freezes as largely pure ice) i.e.:


So the water activity is a function of the temperature alone (see Table 5) and is not influenced by the food composition and is of no use in understanding how well food is preserved.  Despite this limitation the aw concept is widely used in dried, fresh, and salted foods.


Moisture sorption

We now have a parameter which describes how actively water will play a part in reactions.  We would also like to get some better understanding of how this parameter responds to the environment; "How will my food get more stable as I dry it?".  Different foods will respond to drying in different ways.  This achieved using a moisture sorption isotherm and is measured as the relationship between moisture content and water activity.  A typical moisture sorption isotherm is shown in Figure 18.   As noted previously, the vast majority of water in a food is bulk water, effectively the same as a dilute solution.   This water is quickly lost as water activity decreases (another way of looking at this would be you can dry almost all of the water from the food without really increasing the product stability.  The more interesting parts of the sorption isotherm are at the bottom of Figure 18 and these are revealed by reploting the data (see Figure 21 or 22 - note the shape and ignore the additional details).

Typical shapes

The two main shapes found in food isotherms are sigmoidal and J-shaped.   Most foods are roughly sigmoidal, particularly if the have a significant polymeric component.   Some sugar based foods such as confections show a J-shaped isotherm.

wpe4.jpg (27494 bytes)

We will be most concerned with sigmoidal isotherms.


Zones of isotherms

See Figure 23f.   We can use the isotherm to infer some details of how the water is bound to a food. 

Note that an individual water molecule will rapidly and continuously exchange between the three layers.  The boundary between zones 1 and 2 is the critical moisture content that must be achieved for maximal stability.


The BET monolayer

The Zone 1-Zone 2 boundary is hard to measure directly on a real isotherm as the curves tend to be very broad and the point of inflection is poorly defined.  Some better understanding of the monolayer value   can be found by fitting a theoretical equation to the measured data and using the theory to calculate the monolayer value. 

The original isotherm developed by Brunauer, Emmett, and Teller (1938), the BET isotherm, was developed to look at the adhesion of gas to a porous surface.   They assumed two types of sorption:

Using a variety of theory these scientists were able to come up with a general equation relating the equilibrium moisture content (m) with the water activity, and two constants - one of which is the monolayer value m0; i.e. m=f(aw,constant, m0).  A manipulation of the equation is shown as Equation 8 and Figure 24 provides a practical example of the calculation of the monolayer value.

There are very many of these equations with more or less strong theoretical basis.  one of the most widely used is the Guggenhein-Andersen-deVreis (GAB) isotherm:

m is the moisture content,

Three empirical constants:
m0 is the monolayer content,
mm is the multilayer constant,
c is the temperature constant


Food spoilage

Figure 23 shows the relative rate of a range of deterioration reactions common to foods as a function of moisture water activity.    With the exception of lipid oxidation, all of the rates decrease at least 100 fold as the zone 2 water is removed and effectively stop at the monolayer value.   This is because whatever the reagents responsible for a reaction, they always need a solvent to move around in in order to encounter each other and react.  As the solvent is removed the rate decreases and, as monolayer water is not adequately liquid-like to act as a solvent, the reaction stops.

Several rates may slightly decrease at high water activities due to dilution of the reagents.

Lipid oxidation shows the same decrease in rate on drying for similar reasons but increases as water is removed below the monolayer value.    The reasons for this are not particularly clear, but are believed to include decreased hydrolysis of a crucial oxidation intermediate and the dehydration of a metal ion catalyst.

Microbiological growth is governed by similar effects as aqueous chemical reactions for similar reasons.   Most organisms have a critical limit below which they cannot grow (Table 6) but the validity of the moisture sorption approach is called into question as the precise value depends on the material used to achieve the water activity (Figure 15). 

Sometimes physical deterioration can also be linked to changes around the monolayer value, for example a crunchy cereal may soften or dried fruit may become too chewy.


Multicomponent foods

Different food materials with different water activities and moisture contents are often packed together as composite foods, e.g., cheese and bread in pizza, salad in burgers, dried cereal with raisins.   The moisture sorption approach is a useful way of determining the migration of water between the components.

The driving force for diffusion of water is differences in aw.  Water will move from a high to a low water activity region if a path is available for it.  

wpe1.jpg (27901 bytes)

In the figure imagine two different foods with different moisture sorption isotherms.   They are initially (position 1) dried to the same moisture content then packed together.   Slowly moisture will diffuse from the high to the low moisture activity until equilibrium is reached (note this need not occur at half the distance between positions 1) at position 2.   The rate of reactions in the two components may have drastically changed and foods which were independently stable may become unstable when packed together.  This analysis gives a good indication of what should happen thermodynamically but no information of how quickly the diffusion may occur.

If you need to pack moist and dry foods together it is often useful to reduce the daw by adding a humectants to reduce the water activity of the moist component.   An ideal humectant would be highly soluble, significantly decrease  water activity, and crucially have no flavor.   No such ideal compounds are available but salt, sugar and sugar alcohols (esp. glycerol) are widely used.   It is also possible to place a barrier between the components, an edible film or a coating of fat may be suitable.