Colloidal gels from proteins and peptides - Physical Gels from Biological and Synthetic Polymers - page 264

this diagram the possibility of gel formation for any given concentration and temperature
can be predicted. Results are shown here for both

β -Lg and BSA; the latter lies below and

to the left, so the system gels more readily.
Experimental studies on this type of state diagram have been carried out on other
systems, using a variety of techniques, by Kawanishi and Tan for synthetic polymers (Tan
et al., 1983 ; Kawanishi et al., 1986 ) and by Tanaka for biopolymers (Tanaka et al., 1979 ).
In most of these diagrams, the two-phase region, as well as the gel region, was located
on the lower-temperature side, i.e. a two-phase region exists at lower temperatures,
giving an upper critical solution temperature (UCST) diagram. By contrast, when the
two-phase region is located on the higher-temperature side, as here, we have a lower
critical solution temperature (LCST) phase diagram. Work by San Biagio and co-workers
(San Biagio et al., 1996 ) suggests that complex LCST type phase behaviour (there
attributed to spinodal demixing) is seen for unfolded BSA, but does not appear to be
present with the native protein.
Using small-deformation oscillatory measurements, Kohyama and Nishinari ( 1993 )
compared the gelation process of 11S and 7S soy proteins in the presence of the common
acidi er glucono-delta-lactone (GDL). The gelation rate was greater and the gelation

time shorter for 11S than for 7S. Nagano et al.( 1994a ) made small-deformation measure-
ments on soy proteins and found that, by plotting log G 0 for glycinin (80°C) and
β

-conglycinin (65°C) against the log of FTIR absorption at 1618 cm − 1 from their work
described above, a very good correlation was seen.
Kohyama and Nishinari ( 1993 ) also examined the effect of GDL on large-deformation
uniaxial compression of gels as a function of time and found a tendency similar to their
small-deformation results. There are very few other reliable measurements in the
large-deformation regime, except those from Foegeding and co-workers (Foegeding,
2006 ). More recently it has become possible to study (small amounts of) protein gels
during fracture under the microscope, especially using the environmental confocal
scanning method (Olsson et al., 2002 ). By using a tension compression stage, changes

in stress and strain can be correlated with changes in microstructure, as demonstrated by
Plucknett and co-workers (Plucknett et al., 2001 ). Using essentially this method, Ohgren
et al.( 2004 ) were able to demonstrate that the fracture behaviour of particulate

-

β -Lg gels

changed with protein concentration. A gel showed brittle behaviour when the

clusters were rigid, and the crack propagated smoothly compared to a gel with an open
network structure, which showed discontinuous crack growth. Differences in the exten-
sibility of the aggregated

' dense '

-Lg structure, induced by addition of a polysaccharide
(amylopectin) solution, were shown and these were related to differences in stress

β

- strain

behaviour and crack propagation.
While all of the above involve heat-set gels, a recent innovation in this area is the
discovery and characterization of globular protein gels that form on cooling. In effect,
solutions are heated at concentrations below the critical value and then Ca 2+ (or
presumably other divalent cations) added. The work by Bolder and co-workers
(Bolder et al., 2006 ) has characterized a number of these, including from

-Lg and
from mixtures of this with other whey proteins. Other work has examined the same
effect for soy protein gels (Maltais et al., 2005 ). At one level there is no great surprise in

β