Chris The Great - 16-9-2006 at 15:35
For some time I have been trying to make boron halides OTC. The process is simple on paper, and I nearly have it all worked out in real life- except
the damn reaction vessel! However, the reactor I envision would work for far more than just this one reaction. It would work wonders for phosphorus
halides as well 
Here is a basic diagram:
The reactor is on the left. It is a 6 inch section of steel pipe, 1" inside diameter, threaded on both ends. The caps are steel end caps. The one
of the left has a hole drilled into it, and the 1/2" copper pipe goes through it- it's a bit smaller than the pipe though so I "reduced" the pipe in a
careful fashion with a hammer. The end of the pipe, on the inside of the reactor, is spread open, so it's wider so the pipe can't get pulled out.
The tube on the inside is 1/4" copper pipe, nad extends to the end to let the chlorine into the reaction mix. Then, the gas travels up the 1/2"
copper pipe and out the pipe labeled "out".
Now, the connection between the 1/2" pipe and end cap is sealed with "stove and gasket cement" this neat stuff which is supposed to handle up to
1650*C. The problem is that it does not, it starts cracking and chipping, and therefore it leaks!
Everything else in my setup worked, chlorine didn't kill me (I usually have bad luck with the stuff), the reactor pipe was glowing bright orange on a
sunny day, but the huge green flames coming out from that joint clued me into the fact my BCl3 was coming out in the wrong place. I mean, it was
spectacular to see brilliant green and bright blue flames leaping up, but it was not as spectacular as boron trichloride dripping out of the condenser
into the waiting methylene chloride would have been.
Heat was provided by a stream of propane gas in the open. No furnace, just put the pipe in a fire pit and hit it with the gas.
So, this is my question: does anyone have any tips on how to improve this part of my reactor? Everything else works as it should, but this part
always cracks leaks apart from the heat.
My cement is not working, but has anyone used a specific brand and found it to be good under this conditions? Should I replace the copper with steel
and have a friend weld it?
Thanks!
solo - 16-9-2006 at 15:45
How about some high temperature silicone gasket material gel, it usually can widthstand some high temperature............something that just came to
me .......solo
The_Davster - 16-9-2006 at 16:23
Another alternative would be to use a longer pipe so you only heat the end with the reactants. Then the seal could be made of a less heat sesitive
material.
Also, you have read of the method in another thread here in which salt and B2O3 are heated to a high temp and BCl3 is given off, yes?
https://sciencemadness.org/talk/viewthread.php?tid=6641
IrC - 16-9-2006 at 18:15
You are dealing with two very different expansion rates, one metal, one ceramic compound. No way to cure given what you have here, it is always going
to break. Making metal ID to OD seals very precise and tight would work well assuming the metals are all the same or at least expand at the same rate.
Also you would then not need the ceramic cement.
Eclectic - 16-9-2006 at 19:12
Use a U-shaped reaction tube: Gas in one end, product out the other. Connections can be screwed, brazed, or welded into the pipe caps.
12AX7 - 16-9-2006 at 19:34
And speaking of brazing, you might put a dab of borax (or other flux) followed by brass (filler) onto that joint. It sounds like you can get it hot
enough.
Tim
Chris The Great - 17-9-2006 at 00:50
Thanks for the suggestions. I had not read about using salt, but I think using salt would require more propane than I would like to use- I am not
sure of the temperature's I'm getting but it is glowing a nice orange on a sunny day. Someone with more experience?
I could probably get it hotter if I put some bricks around it to keep the heat in and the flame around it, I noticed the bottom was hotter than the
top (bottom= bright orange, top = dull red). Probably because the top was radiating out heat to make my leather gloves smoke from a foot away 
So does my setup get hot enough to try salt? Chlorine only requires it to be heated to 450 degrees for the reaction to start compared to 800 for
salt...
I like rogue's idea, assuming I can find a long enough pipe. The local stores don't seem to hold very big threaded steel stuff... I'll check and see
if I can find a 2 foot section, hopefully it doesn't cost too much... It seems like the best idea actually, mainly because it's simple. Simple is
very good for me, at the moment I am only home on weekends due to now being at university.
I could probably just seal it with a rubber sealant (stuff works wonders!) and wrap and wet cloth around the end to keep it cool... but I found that
even the copper pipe did not conduct heat much once out of the fiery inferno surrounding the reaction vessel. I found it only warm to the touch a
foot from the edge of the fire.
I'm not sure what you're describing Eclectic... Could you give me a more detailed description please?
not_important - 17-9-2006 at 01:07
Orange is roughly 1000 to 1100 C, cherry-red is upper 800s to upper 900s, if that is what you are asking.
I believe that Eclectic is suugesting that you create a U-shaped tube, mount it just like the letter U, heat the bottom portion; the upper arms would
stay cool enough that more conventional gasketing or sealent could be used on them.
Note that halogens and boron halides react with many organics, true rubber being one of them. Silicones might be a better choice.
Chris The Great - 17-9-2006 at 01:35
Ahh, ok, well I am sure I have silicon sealant somewhere too. I suggested rubber because that is what was found in the chaulking gun. LOL
Yes, colour = temperature is what I was asking, thanks. I'm quite impressed now that I managed to heat it that hot! I'll try adding some bricks and
see if I can get the whole thing to be orange. This is fun, I wish I had the time to make a proper furnace! Instead, quantity over quality- hit it
with enough propane and you'll get it hot enough LOL
I get the U idea now too. I don't think it would be as effective, as right now the tupe ensures that the chlorine bubbles through the molten mixture,
while in a U shape this would not work or I'd need a really long tube. I think I'll stick with the single pipe just because it's more simple to
construct.
not_important - 17-9-2006 at 07:29
Well, what you have is in effect a concentric U-tube, just that a one-piece setup would avoid the joint problem. with a true U shape it would still
bubble through the reaction mix, just that it would be displacing the mass - you're forcing gas down one arm and letting it flow out the other.
Note that you are heating part of the rig to near the melting point of copper. An all steel, or using ceramic or silica for the inner tubing, might be
better if you really want to take it that hot.
If I was to do this, I think I would go for all steel. Make the wider section somewhat longer, so as to keep the reduction joint in the center
somewhat cooler. Use a T-joint for the 'out' connection and side-arm, perhaps transitioning to copper to get some flexibility in leading away the
product. The inlet tube would use another reduction joint, you could either use 2 tubes - one from each side, or ream out the threads on the smaller
part of the reducer and just seal with silicone or teflon tape. Stand the outer tube with endcap vertical and use a thermit charge to put some molten
iron into the cap end to weld and seal it. This does mean that all access would have to be done by disassembling at the reduction joint.
Another possiblity would be to have a longer tube and heat a section somewhat above the end cap, with the center tube not reaching as deep but just
past the area you heat. Run the rig through sheet metal shield that will sit between the end cap and the heated section. This way the end cap might
stay cool enough, or could be cooled with an air blast, and the reaction mix freezing near the end would seal it.
You are just doing boric acid/boron oxide and chlorine? And would try a phosphate for PCl3/POCl3 ?
tr41414 - 17-9-2006 at 09:44
I couldn't find the patent 832,096 LONDON on B2O3+NaCl reaction, but maybe substitution of NaCl with lower melting chloride salt/mix would do the
trick to bring the temperature needed for the reaction to proceed down to ~450 deg. celsia(mp of B2O3)...
The setup for this rxn can be a closed one, with just one pipe leading out. Pipe, if long enough can also be used as a pre-cooler or even condenser
for the product.
Edit: Sorry for me not making enough effort in searching for the patent. According to the patent temps 900-1000 work and there is no mention of the
non-alkali metal, so that is probably not possible (but why not try it?)...
[Edited on 17-9-2006 by tr41414]
Eclectic - 17-9-2006 at 10:23
The patent:
http://v3.espacenet.com/results?DB=EPODOC&sf=n&NUM=g...
Chris: I didn't realize your reactant was molten, I assumed you were chlorinating boron powder or AlB12 crystals as you did not give the reaction.
Yes, a U tube reactor would not work very well unless the reactants were solid and granular to allow the gas to flow through.
Chris The Great - 17-9-2006 at 16:16
Ok, well now I understand why you though of a U-tube. It would work fine if the reaction was solid.
not_important: I will be trying for POCl3 once I get BCl3 to work, since the reaction is essentially the same in precursors and temperatures.
B2O3 + C + Cl2 -> BCl3
P2O5 + C + Cl2 -> POCl3
I will also replace the inner tube with steel, to prevent it from melting. I will keep the reduced section copper merely since I alreay have a lot of
it and I don't like to spend lots of money.
Eclectic - 17-9-2006 at 17:41
There is a prep in "Kings Survival Chemistry" for PCl3 that involves
firing superphoshate fertilizer with charcoal to make calcium phosphide, then chlorinating that to make PCl3.
Most of the technique in Kings looks flawed and potentially deadly to me, but that reaction might have some promise.