Would a Dewar condenser with dry ice and acetone work for what you need?

Talks about pumps, solvents, and temperature. http://127.0.0.1/talk/index.php/topic,508.0.html

Idea #1:
i found this hive post at "hiveboard/newbee/000494095.html" which brought up a good point:
A solution of dry ice/IPA (much safer than dry ice/acetone) would give a solution of ~ -78C. The reflux temp of an Al/Hg would be 65-70C using MeOH as the solvent. Pyrex has a thermal shock rating of ~160C and the temp difference mentioned above is ~150C so you're really getting pretty close to the danger zone of running a risk of thermal shock. Perhaps using the reaction flask on the bottom, then stack my typical allihn condenser fed with cold water attached to that reaction flask, then put a dewar condenser attached to the top of the allihn condenser. This way the allihn condenser in the middle would act as a buffer to at least slightly cool the vapors some before they reach the dewar.

my only worry about this is that the subzero MeAm dripping back into the nitro Al/Hg might be so cold that it could cause the reaction to come to a halt. Do you think this is a worry or that the nitro al/hg runs so hot that it would simply heat that subzero MeOH/MeAm condensate in no time? Rhodium once stated that if using freezer-cold ketone it would slow imine formation and give a competing reaction - the formation of the alcohol - more of an advantage to occur. I've seen no references or other claims for this aside from Rhodium's one comment. But there is also the worry that such cold condensate being refluxed down is so cold that it might be equivalent of putting the flask in an ice bath (or am i really overestimating the impact of such cold condensate refluxing back down?), which from what i've read is a big no-no and can cause the reaction to come to a halt. Does anyone think this would be a real concern or is the nitro al/hg hot enough to handle the subzero condensate?

Option #2:
If the subzero condensate is indeed too cold and might kill the al/hg or slow imine formation considerably then perhaps instead of using a dewar at all with a dry ice-chilled coolant, abandon that idea for using a longer condenser (allihn or friedrich?) and feed that reflux condenser with subzero salt water that is being recirculated from a bucket of saturated NaCl water solution that has an equal amount of crushed ice in hopes of reaching -10C to -20C range in temps. That should be good enough to condense some MeAm (bp = -6C) or at least increase its solubility in the MeOH condensate considerably due to the lower temp. I read that CaCl2 + H2O can reach even colder than NaCl + H2O and would be a happy medium, however i think CaCl2 might very well kill the water recirculating pump, so if CaCl2 + H2O were employed then it should probably be best kept in a dewar like Option #1.

Here's where I got the info on the NaCl + H2O and the CaCl2 + H2O temps:
h**p://www2.uni-siegen.de/~pci/versuche/english/v105-2.html

i know it doesn't take much to condense the MeOH, but condensed semi-hot MeOH doesn't hold MeAm very well so the MeAm still escapes, it's the MeAm i'm concerned with. so far it seems like the best way to go would be a nice, long friedrich condenser fed with ice cold salt water or ice cold CaCl2 water, if the CaCl2 doesn't clog up / destroy the water recirculating pump. is a friedrich condenser really going to be more efficient for this purpose than an allihn? i only have experience with allihn and liebig condensers

=================================================
the bottom line:
sorry for this long post but i'm brainstorming out loud. here is an excellent SM post (h**p://www.sciencemadness.org/talk/viewthread.php?tid=10444) about freezing different mixtures, the CaCl2 + H2O mixture gets some recognition here for possibily being able to condense NH3 and MeAm, exactly what i'm looking for. just not sure if a plastic water recirculating pump can handle a super cold CaCl2/H2O solution to pump through a reflux condenser or if that type of solution is more of a job for a dewar

why not run the variation where you reduce the nitro methane first?
to avoid all the exotherm from two reactions taking place at the same time?
also one could also try antifreeze i read it's gentle on pumps.
just add dry ice to the antifreeze to keep it's temperature in range.

thanks for the input jon, you're referring to abacus' method?  i've thought of giving that a try but then thought if i go that far, might as well just make one's own MeAm.HCl and do it Osmium style.  i guess there would be a benefit to Abacus' method though because i would think reducing nitro would make a much cleaner product than making MeAm through other routes.  the only problem is that i've heard some bees say a 5-6% solution of MeAm is the best you're going to get in MeOH... or is that number based on the concentration after distillation of the meoh to remove it form the al sludge?

Abacus's method is your best option if you have nitromethane since you won't have to purify it