I think we should start a new thread on peltiers and their uses, but to answer your question about fractional distilation...
They work quite well for distilation of low boiling substances, I have done this with a very quickly rigged setup. I didn'nt continue with the device however because it was unlikely to be able to handle distilations requiring 130 deg Celcius. Most peltiers are destroyed by temps as low as 70 deg Celcius. Marlow has a couple of devices rated for high heat, here is a link to their web site.
http://www.marlow.com/index.htm
They discuss PWM control and have extensive informtion on matching specific devices against heat load.
In the distilations I was doing, I was using the Peltier to heat the flask, instead of cooling it. This would all depend on what you were distilling of course. What is nice about Peltiers is you can switch from heating to cooling by reversing the direction of current flow. So you could transition from 50 deg Celcius above room temperature to 50 deg below all with a single stage device and a heat sink.
If you push the device to hard however, small fractures can develop between the small internall elements stopping current flow and killing the device. I have done this to several devices.
By stacking a smaller device on top of a larger one, you can increase the temperatur difference you can achieve. So if your first stage device drops the temp by 50-60 deg Celcius, a second one could start there and shoot for another 50-60. The efficiency does however drop off.
What makes Peltiers somewhat TrickE, is the parasitic heat produced interal to the device. The device 'pumps' heat from one side to the other side, but it also adds waste heat as a byproduct which must be ejected or removed by the heat exchanger. If heat is not removed at the required rate, the chip will quickly exceed its operating temperature and destroy itself. This can happen in 60 seconds, I have done it myself.
The key with Peltiers is heat ejection, if you can keep the 'hot' side cool, the cold side will be, as much as 70 deg cooler. So if you are cooling the 'hot' side with ice/MeOH to -20 deg Celcius, you will easily get -60 deg Celcius on the cold side even under a 50 watt heat load. But your cooling system will need to remove maybe 80 watts of heat from the hot side of the peltier, 50 that is being pumped and 30 or so added by the device itself.
To exchange that amount of heat and still remain at -20 deg, the coolant will have to be at least 20 deg colder. I need to site down and calculate the exact numbers again, but I have sketched it out several times before. The difference in temp and the conductivity and surface area determine the amount of BTU's per second which limits everything else.
When you over drive your Peltier and pump more heat than can be removed from the device, the 'hot' side gradually rises and the cool side follows it. Before long, seconds in some cases, the 'cool' side can go above room temperature becomming a heater and soon after becomming trash.
Meaning that a poorly cooled device might produce a maximum cooling result of only 30 deg below room temp when powered by 25% of its full power rating. And if driven by 50% of its full power rating cease to cool at all and actually start to heat the 'cool' side above room temperature.
That is why you need tempurature sensors on the hot and cold sides of the device to adjust the supply current dynamically. A simple control program can make the necessary adjustments to account for variations of heat extracted from the reaction, variations in the coolant temperature and changes in the desired target temperature for things like extended crystalizations or distilations.
TrickE
And on the eight day, God created Meth...
... and hasn't done much of anything usefull since!