Sciencemadness Discussion Board

Ampouling Gases

elementcollector1 - 21-1-2013 at 21:17

I recently noticed that every ampoule I've ever made... sucks. They all leak. These are all the glass pipettes (as per NurdRage) sealed together on both ends, and the only one I've ever gotten right was for bromine. So, what's the best way to ampoule gases?

Also, an additional note on elemental gases: Are they miscible with air? That is, if one had a sealed tube filled with air, and they bubbled hydrogen into it from underwater, would the tube be slowly filled top-down with hydrogen (because hydrogen is the lightest gas, and insoluble in water) or would the glass tube now contain a mix of hydrogen and air? If so, what would be a way around this for most gases, if not all?

woelen - 22-1-2013 at 00:49

Gases are miscible, always! If you want to ampoule gases, then you have to accept some impurity of the gas. I succesfully ampouled Cl2 and NO2 years ago and these ampoules still are filled with the gas.

First of all, you need to use real ampoules, with a flat bottom and a narrow top. Fill the ampoule with the gas. Do this copiously, use a lot of excess gas to drive out all air. Then quickly heat the narrow top of the ampoule and seal it. I heat the ampoule until it is red hot and molten and then take a metal clamp with which I quickly seal the top and then I keep on heating for a while to make the seal smooth. In this way I could make samples of probably better than 90% purity. See the wiki page on chlorine, it shows my sample of chlorine gas: http://en.wikipedia.org/wiki/Chlorine

You also have to make the gas sample really dry, otherwise you will have condensed water on the inside of the ampoule when it cools down a little bit and that looks really ugly. I use a mix of 90% CaCl2 and 10% P4O10 for drying the gas.
I only ampouled the visible gases Cl2 and NO2. I see no reason why I should ampoule H2, O2 or any other colorless gas. These samples are not interesting at all and then I also could ampoule air. If you want a nice sample of invisible gases, then try to find discharge tubes containign the gas and show the light of the gas discharge when a high voltage is applied across the discharge tube.

phlogiston - 22-1-2013 at 01:58

Do you use a cold or hot metal clamp to seal the top?
I find that if I use a hot clamp, it sticks to the glass, but if I use a cold clamp, the glass is not well sealed.
This does improve if I heat the seal for a bit, as you describe, but when the ampule contains a volatile liquid I am always worried that the heating may cause excessive overpressure.
Do you ever cool the bottom part with the content while heating the seal?

[Edited on 22-1-2013 by phlogiston]

[Edited on 22-1-2013 by phlogiston]

woelen - 22-1-2013 at 02:46

I use a somewhat hot clamp. I heat the clamp in the flame for a few seconds before I clamp the glass, but it is not red hot. Use of two flames can be convenient, one flame for the ampoule and the other for the clamp.

When the ampoule contains a volatile liquid, then you do not need to worry about strong overpressure. You just heat the top and if all is well, then you fill the ampoule, such that the liquid does not extend into the neck of the ampoule. For display purposes I only fill the ampoule with a small amount of liquid (e.g. sample of Br2 contains 1 to 2 ml of liquid bromine in a 10 ml ampoule), for storage purposes I fill the ampoule 75% to 80% (thionyl chloride, 17 or so ml of thionyl chloride in a 20 ml ampoule).
It is important, that after the seal is made with the clamp, that only the upper rim is heated and not the rest of the neck of the ampoule! I just let the roaring flame touch on the rim, such that it melts. The only place where it can leak is at the rim, so there is no need to heat other places.

I never cool the bottom of the ampoule while sealing it. I keep it in my hand and that would not be possible if I had to cool it. I never had a cracking ampoule, due to heating. Of course I also have failures, but these always are near the neck (e.g. cracking of the sealed part when it cools down). The failure rate is appr. 20% for me.

[Edited on 22-1-13 by woelen]

watson.fawkes - 22-1-2013 at 05:28

There's a wide variety of graphite manipulation tools for working glass, including different kinds of tongs or tweezers. Also, there are formulated coatings (some containing graphite, as I recall) for metal tools that contact hot glass that prevent or minimize sticking.

The main contaminant from bubbling hydrogen through water is going to be water vapor. It's common enough to have acquired its own name, "wet hydrogen", as opposed to dry hydrogen, which has been dessicated (and usually also deoxygenated).

MrHomeScientist - 22-1-2013 at 10:25

Woelen,

Have you ever tried ampouling cryogenic gases? Specifically, I want to liquify some chlorine and ampoule a bit of the liquid, so that when it warms up and evaporates I get a nice concentrated sample of the gas. That will make it easily visible for display purposes. Do you think that's possible, given the large temperature difference between the top and bottom of the ampoule? I'm worried about the glass shattering. I don't need to store the liquid itself (though that would be really cool, if the ampoule can take the pressure), I just want a sample for my element collection that is easy to see.

elementcollector1 - 22-1-2013 at 10:34

Would "Dry-Z-Air" (basically calcium chloride) work for drying hydrogen?
So, could I clamp the pipette-ampoules closed, then?
Unfortunately, I don't have a blowtorch that stays lit, I have to constantly press the button. So, I'm thinking of using one hand to hold the torch, the other to hold the clamp (twist to seal, right?), and clamping the bottom (or top, in the case of hydrogen) to a lab stand.

Lambda-Eyde - 22-1-2013 at 10:39

How do you plan on flame-sealing the ampoule without igniting the hydrogen?

Endimion17 - 22-1-2013 at 11:04

MrHomeScientist , don't even try that. It's way too dangerous. It requires above average ampuling skills and tiny ampoules with heavy walls. Probably an oxyacetylene torch, too, because ordinary torches are not only too cool, but don't give enough heat. Quartz tubes are a better choice, but their softening range is very high.

Finishing pressurized ampoules (and the pressures we're talking about are substantially higher than 1 atm) have to be flawless. Pinching with cold metallic tweezers introduces stress, fucks up the geometry of the seal and you basically end up with a poisonous bomb in your hand that can detonate any time and release a great deal of poison.
The seal must be spherical and cooled slowly.

Chlorine, sulphur hexafluoride and xenon are good candidates. I doubt CO2 is. The pressures involved are too high.

It's best to encase such ampoule in a block of transparent resin, for extra protection.

[Edited on 22-1-2013 by Endimion17]

MrHomeScientist - 22-1-2013 at 11:14

Endimion17:

There was some discussion about this a while back in <a href="http://www.sciencemadness.org/talk/viewthread.php?tid=22473#pid267286">this thread</a>.

My plan, which I am apparently no good at describing, was to make a small amount of liquid Cl2, add that to an ampoule, let nearly all of it evaporate, then seal the ampoule. In this way, there is no actual liquid remaining in the ampoule when it warms up to room temperature, but the slight pressure it produces after it boils would (I hope) be enough to make a clearly visible sample of the gas.

I don't use clamps to seal my ampoule; rather, I melt the neck closed while pulling the top and bottom apart slowly. It usually takes a few minutes with a propane torch. Certainly not a spherical seal, but not quite as stressed as using (relatively) cold clamps.

Endimion17 - 22-1-2013 at 15:12

Quote: Originally posted by MrHomeScientist  
Endimion17:

There was some discussion about this a while back in <a href="http://www.sciencemadness.org/talk/viewthread.php?tid=22473#pid267286">this thread</a>.

My plan, which I am apparently no good at describing, was to make a small amount of liquid Cl2, add that to an ampoule, let nearly all of it evaporate, then seal the ampoule. In this way, there is no actual liquid remaining in the ampoule when it warms up to room temperature, but the slight pressure it produces after it boils would (I hope) be enough to make a clearly visible sample of the gas.

I don't use clamps to seal my ampoule; rather, I melt the neck closed while pulling the top and bottom apart slowly. It usually takes a few minutes with a propane torch. Certainly not a spherical seal, but not quite as stressed as using (relatively) cold clamps.


Oh, that's a different thing. :)
However it's not neccessary and it can easily lead to acute poisoning. Have you ever tried to ampoule chlorine using a gas generator spewing out chlorine ~0.5 ml/s? Holy crap, it has to be done outside while standing upwind. It's horrible.

First prepare your ampoule by narrowing its neck just to let a tiny tube inside. That way you won't waste time when the sealing starts.
Then flood the ampoule with chlorine. Slowly pull out the tube, so that the air doesn't rush inside. Chlorine should be constantly pouring out while you pull out the tube i.e. there has to be a positive chlorine pressure inside.
After you pull it out, seal it immediately with neck pointed upwards. Only traces of air remain inside.

It would be more interesting (and a lot more difficult) to fill it with pure chlorine and then evacuate it, so that low pressure chlorine remains inside.
Why? To use it in this manner.

watson.fawkes - 22-1-2013 at 19:16

Quote: Originally posted by elementcollector1  
Would "Dry-Z-Air" (basically calcium chloride) work for drying hydrogen?
[...]
Unfortunately, I don't have a blowtorch that stays lit, I have to constantly press the button. So, I'm thinking of using one hand to hold the torch, the other to hold the clamp (twist to seal, right?), and clamping the bottom (or top, in the case of hydrogen) to a lab stand.
A desiccant will make the air more dry. Just how dry do you need it? You haven't said what you want the gas for. Without knowing the application, I can't offer appropriate advice.

Having stated that, if what you want is something for an element collection, then all you really need is for the dew point of the water inside the ampoule to be low enough that you won't ever see condensation. A desiccant should accomplish that much.

watson.fawkes - 22-1-2013 at 19:34

Quote: Originally posted by MrHomeScientist  
My plan, which I am apparently no good at describing, was to make a small amount of liquid Cl2, add that to an ampoule, let nearly all of it evaporate, then seal the ampoule. In this way, there is no actual liquid remaining in the ampoule when it warms up to room temperature, but the slight pressure it produces after it boils would (I hope) be enough to make a clearly visible sample of the gas.
If you can make a T-joint in glass, you can make your own ampoule rig. The trick is to fill the ampoule from a mostly sealed system, and to fill it upside down with vapor from inside the system. Consider a long tube, sealed on both ends, mounted vertically so that the bottom acts as a reservoir, with a T-joint above the reservoir. Fill the reservoir with liquid chlorine (chilled, obviously), seal off the port, turn it upside down, open the port and let some of the chlorine evaporate to purge the vessel (you could also use vacuum technique, if you want higher purity). After purging, there's only the target gas within the system. Close the port and turn it all right-side up again.

Now create an ampoule by parting off the top end of the vessel. Clamp the vessel at the reservoir end in a stand. Clamp the top end with a hand clamp (bend up some wire if you need to). Apply heat to the top tube section at a height right for however large you want your ampoule to be. As the glass softens, twist the top end, pull out a bit and seal the glass to itself. You'll end up with two ampoules, one in your hand and one in the stand. You can reuse the reservoir section by cutting off the top end and sealing on a new ampoule section.

Practice first without chlorine, obviously. This is a level of glassblowing manipulation that used to be common in undergraduate labs. It's nothing like making a ring seal, for example.

woelen - 23-1-2013 at 00:33

I never have ampouled cryogenic gases and I would never do that. You need ampoules with thick walls for that, the ordinary thin walled ampoules most likely easily break due to the inside pressure. Another issue with pressurized thin walled ampoules is that when the seal is still liquid, the glass expands like a balloon and this balloon then pops and you again have a little hole in the ampoule. I think that ampouling pressurized gases is extremely difficult without the proper equipment and the proper ampoules. I accept that this is beyond my abilities.

Btw, a chlorine ampoule can be nicely made, even without pressurizing the gas. The ampoule I show on wikipedia is at normal atmospheric pressure and it nicely shows the color of the chlorine. Use a 20 ml ampoule for this gas, this makes it easy to show the green color.

AndersHoveland - 23-1-2013 at 18:30

If you want to practice, you can try ampouling nitric oxide. Keep checking on it periodically. If it develops a brown color, you know there is an air leak.

woelen - 24-1-2013 at 00:08

Amouling nitric oxide (NO) without getting any brown color from the start is nearly impossible. During the ampouling process you always get some mixing of gas and air and if e.g. 10% of the gas is replaced by air, then you already have a fairly deep brown color in the ampoule.

phlogiston - 24-1-2013 at 02:52

Woelen, I've experimented a bit over the past two days with your method of sealing, using pasteur pipets which I first sealed on the other end for practice, and in my hands the flattened piece of glass obtained by clamping with a moderately hot pair of tweezers nearly always (>90%) developed a crack upon cooling. The crack is always in the flat part, but very near the tubular section and it often results in leaking.

After clamping, I heat the end, including the flat bit of glass for another 5 seconds or so, turning it as I heat it, and then place it upright in a rack to cool.

I did have some success with another method, which is (1) heat the end of the tube untill soft, (2) grap the very end with a pair of warm tweezers (3) remove from the flame and pull the tweezers about 10 cms away while twisting, thereby closing the tube and drawing a wire. (4) heat the wire in the middle, thus separating the glob of glass in the tweezers and (5) keep heating the end of the wire, so that it eventually melts completely and forms a little ball of glass at the end of the pipet/ampoule.
It's still not perfect though, with about 30% of the seals failing, especially if I fill the improvised ampoule to a large fraction (say to 2 cms of the end) with a moderately volatile liquid (ethanol).

[Edited on 24-1-2013 by phlogiston]

woelen - 24-1-2013 at 03:26

Could it be that the pasteur pipettes are made of another type of glass? I use real ampoules (see the wiki page for chlorine) and these do not break on cooling down. I also do not turn the ampoule around, I simply firmly press the molten neck of the ampoule between the clamp, such that the glass is pushed flat. I then keep the flat part of the glass in a flame to make a nice rounded rim, otherwise it becomes ragged.

elementcollector1 - 24-1-2013 at 13:22

What if you heated the glass, sealed it, put a clamp on the top and flattened it out, and while still molten, twisted the glass around to form an even (but rather pointy!) neck, then applying a bit more flame to the end to round it out?

Not sure if I described it right, but that's what I see in my head.

AndersHoveland - 24-1-2013 at 14:13

Quote: Originally posted by woelen  
During the ampouling process you always get some mixing of gas and air

In the light bulb industry they use an inert atmosphere (argon and nitrogen) and use a plasma torch, for sealing the fused quartz capsules in halogen bulbs.

Lambda-Eyde - 24-1-2013 at 16:04

Quote: Originally posted by woelen  
Could it be that the pasteur pipettes are made of another type of glass?

I'm willing to bet that these are made from soda glass and that that is the source of the problem. I use borosilicate test tubes for my "ampoules", and they work great with some practice.

watson.fawkes - 24-1-2013 at 16:45

Quote: Originally posted by Lambda-Eyde  
I'm willing to bet that these are made from soda glass and that that is the source of the problem. I use borosilicate test tubes for my "ampoules", and they work great with some practice.
I would concur. The thermal coefficient of soda-lime glass is approximately three times that of borosilicate. It's well-known that soda-lime glass is more fragile when used (see all the stories of USA-brand Pyrex), but it's even more important when fabricating glassware. The thermal gradients are even more severe.

I'm guessing that the technique absent from the folks who are having trouble here is flame annealing. The goal is to cool hot glass slowly. Somewhat counter-intuitively, that means continuing to heat the glass after working it. Yes, more heat, but with a cooler flame. The flame you want is not the neutral flame for working the glass, but a reducing (sooty) flame with less oxygen. Some of the fancier torches and gas setups I've seen have a valve to mix oxygen with regular air. This cools the flame in two ways. Most obviously, there's less oxygen, so the fuel burns over a wider area. Less obviously, it takes energy to heat up the inert nitrogen component, directly cooling the flame. Play the heated area in and out of the flame, leaving it out longer each time.

Edit: typos, clarification.

[Edited on 2013-1-25 by watson.fawkes]