been predicted to be less than the effect of holding at any given
temperature for 3 min [20]. Given that the viscosity at the glass
transition temperature is on the order of 10
13 times the viscosity
under ordinary conditions, this equation predicts very long storage
times, even above the glass transition temperature. For example,
using curve fit parameters for the M22 vitrification solution [23], a
storage time of ~100,000 years is predicted even at À120
C, or
around 4
C above T G [20]. However, there has as yet been no
empirical testing of such predictions, and such predictions will
eventually have to factor in the effects of nucleation above T G
[89, 217]. On the other hand, according to one Scheme [131],
intensive nucleation above T G might not be a problem for the
survival of a living system because nucleation will generate ice
crystals that are too small to be harmful. If the liquids in the sample
are allowed to nucleate completely, the lack of heterogeneity in
crystal size would preclude recrystallization, and warming would
then simply melt the ice nuclei, perhaps without significant grain
growth. The workability of this scheme has never been experimentally tested and might lead to interesting results when tested in the
future.
If storage is to be below T G to minimize nucleation, how far
below T G is cold enough for this purpose? In the M22 vitrification
solution, detectable nucleation can be extrapolated to be extinguished at about À136 to À137
C, or about 13–14
C below
T G [2]. Mehl [217], comparing the devitrification tendency of the
VS41A vitrification solution after 6 months of storage at
À135 Æ 3
C (a mean of 12
C below T G ) to that of unstored
samples, found that the warming rate required to observe zero ice
on warming increased from 50
C/min to only 100–150
C/min
after storage, which supports the idea that extensive nucleation
during holding near T G may not lead to insurmountable problems
on warming. Mehl also pointed out that the number of nuclei may
not matter if rewarming is fast enough to suppress their growth on
warming and they are all the same size at the beginning of warming,
which will tend to be the case since ice crystals nucleated near or
below T G are not likely to grow until warming begins.
Although it may be academic given the above long projected
storage times near T G , structural relaxation times below T G take on
the form of Arrhenius kinetics, rising exponentially rather than
super-exponentially as temperature continues to fall [2]. Essentially,
the extension of logarithmic viscosity plots above T G to temperatures below T G is described approximately by the tangent to the
curve at T G .
From time to time, investigators who correlate the glass transition with the stability of dried systems have noted that deterioration
of those systems can take place even very far below the nominal
glass transition temperature(s) of those systems [383–386]. The
reason for this is unknown, but it must be remembered that the
physical state of dried systems is far different from that of hydrated
80
Gregory M. Fahy and Brian Wowk
temperature for 3 min [20]. Given that the viscosity at the glass
transition temperature is on the order of 10
13 times the viscosity
under ordinary conditions, this equation predicts very long storage
times, even above the glass transition temperature. For example,
using curve fit parameters for the M22 vitrification solution [23], a
storage time of ~100,000 years is predicted even at À120
C, or
around 4
C above T G [20]. However, there has as yet been no
empirical testing of such predictions, and such predictions will
eventually have to factor in the effects of nucleation above T G
[89, 217]. On the other hand, according to one Scheme [131],
intensive nucleation above T G might not be a problem for the
survival of a living system because nucleation will generate ice
crystals that are too small to be harmful. If the liquids in the sample
are allowed to nucleate completely, the lack of heterogeneity in
crystal size would preclude recrystallization, and warming would
then simply melt the ice nuclei, perhaps without significant grain
growth. The workability of this scheme has never been experimentally tested and might lead to interesting results when tested in the
future.
If storage is to be below T G to minimize nucleation, how far
below T G is cold enough for this purpose? In the M22 vitrification
solution, detectable nucleation can be extrapolated to be extinguished at about À136 to À137
C, or about 13–14
C below
T G [2]. Mehl [217], comparing the devitrification tendency of the
VS41A vitrification solution after 6 months of storage at
À135 Æ 3
C (a mean of 12
C below T G ) to that of unstored
samples, found that the warming rate required to observe zero ice
on warming increased from 50
C/min to only 100–150
C/min
after storage, which supports the idea that extensive nucleation
during holding near T G may not lead to insurmountable problems
on warming. Mehl also pointed out that the number of nuclei may
not matter if rewarming is fast enough to suppress their growth on
warming and they are all the same size at the beginning of warming,
which will tend to be the case since ice crystals nucleated near or
below T G are not likely to grow until warming begins.
Although it may be academic given the above long projected
storage times near T G , structural relaxation times below T G take on
the form of Arrhenius kinetics, rising exponentially rather than
super-exponentially as temperature continues to fall [2]. Essentially,
the extension of logarithmic viscosity plots above T G to temperatures below T G is described approximately by the tangent to the
curve at T G .
From time to time, investigators who correlate the glass transition with the stability of dried systems have noted that deterioration
of those systems can take place even very far below the nominal
glass transition temperature(s) of those systems [383–386]. The
reason for this is unknown, but it must be remembered that the
physical state of dried systems is far different from that of hydrated
80
Gregory M. Fahy and Brian Wowk