[labnetwork] Argon in a BCl3 cylinder

[John D Shott]

Sean:

Just to add a bit to the others comments. 

My guess was that, when full, this was something like a 5 or a 9 lb cylinder … or maybe 2.5 or 4 kg equivalent. In other words, many moles …

How much Ar did you add? The head volume of the cylinder at a pressure of 25 PSIG minus the vapor pressure of the BCl3. I’m guessing that is a fraction of a mole. 

Of course, in the vapor it is the partial  pressure of the two components that determines the initial mixture you are extracting. 

However, as you extract gas, argon leaves permanently, whereas the BCl3 is replaced by vaporization. 
So, in principle, it will get more and more pure … but never 100%. By estimating the total flow of gas, you should be able to estimate when you get to 99% BCl3 in the vapor, or 99.9%, or whatever you’d like. 

For example, starting with equal vapor pressures of Ar and BCl3 (that is a 50% mixture), if you extract one “head volume” of gas, you should be down to 25% Ar. One more head volume of flow, you should be down to 12.5%, etc. 

How can you make that happen faster? By chilling the BCl3 so that the Ar is initially the more dominant gas in the mixture and thus preferentially removed. 

At least that is the first order analysis I would consider as an old timer …

Good luck. 

A word of caution: it might be tempting to put the cylinder in a tub of ice water to reduce the BCl3 vapor pressure and, as a result increase Ar extraction. Remember that BCl3 is water reactive and an external leak of BCl3 would react to form HCl. 

Why do I think of that? When I was young … we are talking 1972 or 1973, I think … we had a “Coke syrup canister” full of 50 lbs of SiCl4 for an old epitaxial reactor. To keep the deposition rate constant during a long run, the canister lived in a tub of water to minimize vaporization cooling of the SiCl4. One night the canister developed a leak so that the remaining SiCl4 liquid reacted with the water in the tub and released a whole lot of HCl that corroded everything in sight. Fortunately, this all happened in the middle of the night and we did have a crude scrubber … so nobody was exposed to or injured by HCl. It is, however, not a scenario you want to replicate. 

Have a good day. 

John

Sent from my iPhone

> On Jun 19, 2019, at 4:40 PM, Dwayne Chrusch <Dwayne.Chrusch at umanitoba.ca> wrote:
> 
> Hello,
> 
> If you’re using a 2 stage (or multi-stage) regulator, only the second stage (regulator output) and intermediate stage chambers are contaminated.
> 
> Here’s a link to how a 2 stage regulator works;
> https://www.youtube.com/watch?v=UfaucxS5rIc
> 
> In order to flow gas from the intermediate chamber to the secondary chamber, the secondary pressure must be lower than the intermediate pressure. If the pressure in the secondary is greater than the intermediate, the orifice between the intermediate chamber and the gas cylinder would remain closed. 
> 
> You should be able to purge the contaminated gas out. First, close off the cylinder. Next, use your vacuum system to evacuate the gas from the system. Let the MFC drop to zero flow. Set the gas flow rate to as high as safely possible. This will pull the contaminated gas out of the low side and intermediate chamber of the regulator (assuming it’s dual stage). In theory, there will always be some stubborn Ar gas, but it may be practical to purge “enough”. You could always recharge the gas line, and then re-purge it. 
> 
> 
> Regards,
> 
> Dwayne D. Chrusch, MSc.
> Operations Manager, Training & Safety Coordinator
> Nanosystem Fabrication Lab (NSFL)
> E3-487 Engineering
> 75A Chancellor's Circle
> University of Manitoba
> Winnipeg, Manitoba
> Canada, R3T 5V6
> Tel: +1 (204) 474-8246
> Fax +1 (204) 261-4639
> http://home.cc.umanitoba.ca/~shafaic/NSFL_site/NSFL_Home.html
> 
> 
>> On Jun 19, 2019, at 12:15 PM, Rinehart, Sean <sean.rinehart at yale.edu> wrote:
>> 
>> Hello all,
>> 
>> Recently our facility’s BCl3 cylinder had been charged with argon, increasing the delivery pressure to 25 psi. The cause has been corrected, but now I’m evaluating whether or not the cylinder needs to be replaced, and would appreciate some advice.
>> 
>> Since the root issue was fixed, the pressure has slowly dropped from 25 psi to 15 psi with use. Based on this, it seems like with more use the argon should completely deplete from the cylinder, but maybe someone can tell me if that’s actually true, and if there’s a way to evaluate gas concentrations. Additionally, the plasma glow is whitish instead of pink, so there’s definitely BCl3 content in the delivered gas.
>> 
>> Thank you,
>> 
>> Sean Rinehart
>> Assistant Director, Cleanroom Operations
>> Yale University
>> 203 432 4303 
>> 
>> _______________________________________________
>> labnetwork mailing list
>> labnetwork at mtl.mit.edu
>> https://mtl.mit.edu/mailman/listinfo.cgi/labnetwork
> 
> _______________________________________________
> labnetwork mailing list
> labnetwork at mtl.mit.edu
> https://mtl.mit.edu/mailman/listinfo.cgi/labnetwork