Chemistry Reference
In-Depth Information
adjacent particles, and then aggregation, was shown to take place when
surface viscosity decreases because of an increase in temperature or in water
content, and reaches a critical value depending on the particle size and the
characteristic time of the method used to monitor the changes. The sticky
point (T
s
) and glass transition temperature were observed to be affected
similarly by increasing moisture content, with T
s
close to the T
g end
values,
that is about 208C above T
g onset
(Roos and Karel, 1991b). Caking and
collapse were modelled according to WLF kinetics. For a (T-T
g
) range
between 15 and 308C, the relaxation time for collapse of a freeze-dried
sucrose-raffinose model could be fitted to the WLF expression with the
''universal'' coefficients (Levi and Karel, 1995). Because of the rather narrow
(T-T
g
) range, the C
1
and C
2
values may not be meaningful. What is most
important, however, is the high level of the mean apparent activation energy
(
>
200-400 kJ/mol), which points to the large temperature dependence of the
phenomenon. Caking of a spray-dried fish protein hydrolysate was also
characterized by a WLF relationship, with adjustable C
1
and C
2
coefficients,
for (T-T
g
) between about 20 and 808C (Aguilera and del Valle, 1995).
Recently, a test was developed to monitor the stickiness progression with
time (Paterson et al., 2005). The (T-T
g
) value was confirmed to be the main
factor determining the rate of stickiness development in amorphous lactose,
regardless of which combination of temperature and relative humidity was
used to achieve it. The rate of sticking was fitted to the WLF expression with
adjustable C
1
and C
2
coefficients (respectively, 1.6 and 3.58C), for a (T-T
g
)
range 1-258C (Paterson et al., 2005). Given that the viscosity of various
sugars [e.g. sucrose (Champion et al., 1997a,b), glucose (Williams et al.,
1955)] was observed to show WLF behaviour with coefficients close to the
''universal'' values, we must recognize that the reported values may suggest
viscosity is not the only phenomenon at work in stickiness. Based on this
inconsistency of the reported coefficients, Chen (2007) showed that the
lactose results could be modelled using an Arrhenius-based expression. The
''driving force'' for the stickiness development as a result of glass transition in
amorphous materials would then be (1/T-1/T
g
) rather than (T-T
g
), which
would indicate an independent action of temperature and water content.
For spray-dried milk powders, stickiness and caking are quite differ-
ent depending on their fat content. In skim milk powder (SMP), cohesive-
ness appears to be dependent on the glass-rubber transition of lactose, as a
result of increasing temperature or water content [with a mean activation
energy
270 kJ/mol in the temperature range T
g
+58Cto+158C(Ozkan
et al., 2003)]. In whole milk powder (WMP), the loss of flowability is
observed at a temperature below the glass transition and is related to
melting of fat. As the surface composition of WMP is mainly fats, liquid
fats are supposed to form liquid inter-particle bridges, or to soften the
