CRYO PROTECTANT

[Sean Parkin]

Lori Kohlstaedt (kohlstaedt at sbmm1.ucsb.edu) wrote:

: So I might as well say this since someone is bound to:  I have had the
: experience of having successful freezes for a certain crystal only from
: liquid propane.  Freezing either in the N2 stream or in liquid N2 didn't
: work.  The liquid propane-cooled crystals showed decreased mosaicity while
: the crystals frozen by the other two methods showed greatly increased
: mosaicity.  Maybe this is corellated with crystal size or geomentry?

I have mounted many crystals of many different species, shapes and sizes
both in cold gas streams and by dunking in liquid nitrogen.  Sometimes the
latter will work where the former does not, but I have never experienced 
the reverse.  The difference between gas stream and liquid nitrogen is 
clearly one of cooling rate.  The problem described by Lori Kohlstaedt is
of a different nature though.  Since the crystals were successfully cooled
by liquid propane dunking, it is most likely that the gas stream cooling 
method failed because the cooling rate was not sufficient.  But this does 
not explain the failure of the liquid nitrogen dunk.  However, I can think 
of a few explanations for this behaviour.

I am assuming here that the crystal never gets exposed to anything warmer 
than the cold stream en route from the nitrogen dewar to the 
diffractometer.  If anyone is unsure of how to do this, plans are 
available from either me or Hakon Hope.

I know from experience that optimum antifreeze (cryoprotectant) 
concentrations are not necessarily the same for gas stream and liquid 
nitrogen dunking, so it is likely that optimum concentrations for liquid 
propane are also slightly different.  The reason for this may well be 
something like different contraction rates for the crystal and the 
antifreeze, resulting in stress to the crystal.  If this is the case, 
then the conditions must be tweaked a bit for liquid nitrogen.  

Since the cooling rates in both liquid propane and liquid nitrogen are 
very high, it is extremely unlikely that the actual chemical nature of the
cryogen is important.  The explanation is most likely connected with some 
unforseen physics. Some crystals undergo small nondestructive phase 
transitions.  Due to the large temperature range for liquid propane, it is
quite possible for a crystal to be dunked in the propane at a temperature
just above the transition point, and not undergo any transition.  If the 
cold stream is above the transition point, then the crystal is still 
safe.  On the other hand if the cold stream is just below the transition, 
the crystal will pass through it at a very gentle rate (cooling rate is 
proportional to delta T) and it may get a chance to anneal.  Again the 
crystal will still be ok.  On the other hand, if the crystal is rapidly 
cooled through the transition point to 77K, it will be in a metastable 
state and the resulting strain on the lattice may be enough to cause 
serious mosaicity increase.  Now, if the cold stream is below the 
transition temparature, the crystal will still be in that metastable 
state when it reaches thermal equilibrium, the stress may be alleviated a 
little, but the damage may have already been done.  If the cold stream is 
above the phase transition point, the situation is much worse.  As the 
metastable state (equivalent to the room temperature state) warms, it 
will first pass through the phase transition into the low-temperature 
stable state, and then back again to the state stable at warmer 
temperature.  The damage this could do is easy to imagine.  I have seen a 
situation similar to those described in concanavalin A, which undergoes a 
phase transition at about 165 K.  In addition, the a axis of BPTI is 
longer at 125 K than it is at room temperature - obviously the effect of 
some structural rearrangement.

There is also a dependence on crystal size and shape, and it is a big 
one.  Small crystals are much easier to cool than big ones.

The method of choice for cooling crystals should be the one that maximizes 
convenience, reproducibility, simplicity and safety.  Here nitrogen 
cooling wins hands down for convenience.  Reproducibility is also much 
better than for propane cooling due to the large temperature range of 
liquid propane.  Optimization of cooling conditions for both methods is 
similar apart from the fact that crystals need to be mounted to be 
inspected. Techniques for transfer from liquid nitrogen to the gas stream 
exist that are much simpler than the liquid propane techniques currently 
in vogue.  They also permit safe archiving and transportation of 
crystals.  The temperature during mounting, dismounting, storage and 
re-mounting has been measured and never rises above that of the cold gas 
stream.  If need be, they could be also used with liquid propane.  The safety
considerations concerning liquid propane in open vessels in a laboratory 
are real and obvious.  Liquid propane cooling may have an advantage in 
a few rare cases where unforseen phase transitions exist though.
 
I hope some parts of this are useful to somebody.

Sean