I am trying to tighten up and document the device and instructions that I am using to produce anhydrous NH3, but I have an idea that might simplify one aspect of the system. But I don't want to delay posting the current method unless it looks to make a real difference...
Here is the situation.
The fert and NaOH are added to a reaction chamber dry, which keeps them from reacting to any noticable degree. I then trickle in h2o to trigger the reaction of the strong base stealing from the week base and liberating NH3. However, the h2o absorbes an enourmous volume of NH3. In fact it will reobsorbe all the NH3 produced if given enough time and contact. In some early designs I would initially get a half cup of condensed NH3, which would then dissappear once the NH3 vapor prodution rate fell below the rate of absorbtion.
The rate of absorbtion is loosely determined by vapor consentration, surface area, pressure, concentration of NH3 already in the water, pressure and temp. What I have currently done to cap my NH3 loss is use a couple of one-way-vavles to pull air back into the reaction chamber instead of NH3 when it goes into suckback mode. This does have the unfortunate side affect of 'back-pressure' in the 1/2psi range between the reaction chamber and dryer/precooler/condenser.
This backpressure and the physical restriction of flow increases the pressure in the reaction chamber which increases absorbtion, lowers the presser in the condenser which slows condensation and gernerally slows the flow of the vapor out of the reaction chamber where it is in danger of being eaten instead of quickly hurrying on to the condenser where it will provide drugs for the needy.
Now even with the downsides listed, output with the one way valves is far greater than without, but I am hoping for better if possible.
I have two idea's which might offer better solutions, so maybe the wiser bee's can steer me to the more probabale option or save me two dissapointments.
1) Use a lighter than water solvent layer above the reaction to limit the 'contact' between the h2o and NH3 vapor. In my mind the vapor would bubble through the right solvent and then be blocked from the thirsty hydro. This of course assumes the existance and availability of a solvent that is less dense than h20, does not react with any of the participants and does not absorbe NH3. Does such a solvent exist? And what else should I consider?
2) Flow a stream of dry CO2 throughh the reaction chamber which would increase the flow rate of the NH3 out of the reaction chamber and decrease the consentration of the NH3 in contact with the water. This would also provide a non-condensing gas flowing thru the system making it practical to condense the NH3 inside of a small cross section heat exchanger without worrying about liqued blocking up the flow. The CO2 would propell the dropless through whereas 100% NH3 could just fill up the tube with liquid until the friction and inertia won out.
So this method has additional advantages, but wouldn't seem to provide as complete a solution as the solvent might, if it worked perfectly. But this method does increase the part count to some degree, but since I already use dryice to provide the cooling for the final condenser, I have a handy source for the CO2. Would this work, or am I just shadow boxing?
Finally, I could combin both methods, geting increased flow and reduced NH3/h20 contact and eliminate absobtion losses and gain the advantage of condensing the NH3 within the condenser tube, which is quicker than through the walls of the collection chamber as I am currently doing.
Any advice is appreciated!
TrickE
And on the eight day, God created Meth...
... and hasn't done much of anything usefull since!