Sciencemadness Discussion Board

Replacing air with gas in vessel without vacuum?

silvergrahm - 7-4-2013 at 00:33

Does anyone know of any techniques or devices that can replace air with a gas, down to 99.99%, without vacuuming the vessel first?

Hexavalent - 7-4-2013 at 03:04

It depends what gas, obviously...

for gases "heavier" than air, flooding the container continuously with the desired gas (using an exit bubbler, of course) should be adequate. For those "lighter" than air, I'm not sure this is possible without using vacuum to evacuate the vessel first.

DJF90 - 7-4-2013 at 03:26

Of course, but it depends on how much gas you're willing to expend in such a process. A continual stream of (for example) nitrogen gas will eventually displace all the other gaseous molecules, but the time to do so to the level you strive for could be long. The best way to do it without vacuum is have an inlet and outlet for the gas flow at opposite ends of the apparatus. For example, on a "vacuum distillation" setup, you could have your inlet atop a claisen adaptor or in a side neck of the flask, and the outlet bubbler could be connected to the "vacuum barb" on the receiver bend. This maximises the efficiency of the displacement.

On a related note, even a relatively weak vacuum (say, 500 mbar) can be useful, if one has the right kit. Pumping down to 500mbar (half an atmosphere) removes half of the air molecules. Breaking this vacuum with the desired gas gives you 50%. Repeating this process is a classical example of a geometric progression of an infinite series. The air that remains after n cycles is (0.5)^n, where the 0.5 comes from the ratio of absolute pressure pumped down to, and the atmospheric pressure (500mbar over 1000mbar (approx 1atm) in this case.

If you pull down to 100mbar, then the air remaining after n cycles is (0.1)^n. It can be shown that after a relatively small number of cycles, you can achieve a very high exchange of the air in the apparatus. "But how do you do this repetitive process easily in the lab>", I hear you ask. You can use a double oblique stopcock, or more elegantly, a firestone valve (www.aceglass.com/dpro/attachment_files.php?id=37).

gsd - 7-4-2013 at 04:18

In Chemical Engineering terminology, this is a "residence time distribution" problem.

Depending on the shape of apparatus and method of distribution of fluid, it may take from minimum of 1(for a pipe line) to more than 20 (for a vessel filled with packing) volume changes.
There is very interesting way to determine number of volume change required for small apparatus with irregular or complicated geometry.

Assemble your apparatus and pass water from the lowest entry point and allow it to overflow from highest exit point. Measure the flow rate and keep it steady. Now if volume of you apparatus is say 1 liter then make about 10-15 liter of coloured water in a bucket (few crystals of KMnO4 give nice purple coloured water). Now switch over from flow of normal water to coloured water and measure volume required to get uniform original coloured water at exit.

That is the minimum purge volume you should ensure with your fluid.

Gsd

[Edited on 7-4-2013 by gsd]

BromicAcid - 7-4-2013 at 05:38

Speaking of fluid, never done this before but you could fill your entire apparatus to the brim with an inert solvent then seal and begin purging with inert gas, placing a tube at the lowest point via a torion adapter or the like. Then you're displacing the solvent with your inert gas and you'll be able to see when the only thing left in your apparatus is inert gas and solvent vapor and if you continue to blow it out you'll get rid of even that (not the best of methods but the nose usually knows). Of course then you have to get reagents into the vessel either through an addition funnel or a powder funnel with a constant outflow of inert gas (which gets messy) or needle transfer via septum or any of the other inert transfer techniques. Might be an interesting way to go about it.

[Edited on 4/7/2013 by BromicAcid]

silvergrahm - 7-4-2013 at 18:15

You guys rule. I like the Firestone valve w/ .X vacuum idea best, but all awesome.

Dr.Bob - 8-4-2013 at 07:40

The "inert liquid" trick is neat, you would of course need to insure that the inert liquid was free of all dissolved reactive gases. Plus it is hard to find a true inert liquid. I just flow N2 through the flask a while, that does pretty well, longer for larger flasks, from a minute to 10+ minutes. I like the vacuum/N2 method as well, but not always practical, due to either no vacuum, or very volatile reactants or solvents. I have also learned that too much inert gas can quickly evaporate the solvent or volatile reactants also, if you go too long.