function ofwonly, the relation betweenawandwmay vary considerably.
This then means that the relation between diffusivity andawvaries among
foods. Figure 8.9c shows that the difference can be considerable, even for
intermediate moisture foods. The second point was discussed in Section
4.3.3. The rate constant of a chemical reaction is generally determined by
its activation free energy (DG{), and this also holds for a bimolecular
reaction [Eq. (4.12)]. But ifDG{is very small, orDeffis very small, the
reaction may be diffusion controlled [Eq. (4.14)]. It depends on the
reactions involved and on the further composition of the system at what
(low) water content diffusivity becomes rate limiting. To give a very rough
indication, the reactions in a food are likely to be diffusion controlled if
Defffor water is< 10 ^14 m^2 ?s^1 , unless all of the reacting molecules are
very small.
In many intermediate-moisture foods, most chemical reactions will not
be diffusion limited. In low-moisture foods (e.g.,aw<0.2), however, the
most important cause of chemical stability will generally be small diffusivity.
This would apply to all examples in Figure 8.10, except for oxidation of
carotene. A relation like that in Figure 8.10d is generally observed for
oxidation reactions, where water is not a reactant, and even at aw& 0
perceptible diffusion of O 2 may occur; see further point 6, below.
- Concentration of reactants. For a bimolecular reaction in aqueous
solution, the concentration of reactants increases with decreasing water
content, and the reaction rate would then be proportional to concentration
squared. This is undoubtedly the main cause for the increase in reaction rate
with decreasingawshown in Figure 8.10a; at still loweraw, other factors are
overriding. Also the composition of the reaction mixture may alter when
lowering water content, for instance because a component partly crystallizes
or becomes dissolved in oil (if present). - Activity coefficientsof the reactants generally alter upon water
removal. This was discussed in Section 2.2.5. For ionic species, the activity
coefficient will decrease; for neutral ones it will generally increase, and the
effect can be very large at very lowaw. These changes presumably explain
part of the variation in the relation between reaction rate andaw. Especially
for unimolecular reactions, like the denaturation of proteins, variation in
factors 2 and 3, above, would have no effect, but activity coefficients may
change. This is equivalent to stating that the solvent quality is altered.
Figure 8.11 gives some examples of—what are effectively—heat denatura-
tion rates, and they show great variation. It is even possible that the rate
increases with decreasingw, as for lipoxygenase, despite the notion that
protein denaturation goes along with increased hydration (Section 7.2).
However, even atw¼0.15, water activity would probably be>0.6 at 72 8 C
in the system studied.