So This all started when I landed a huge glove box at a local auction for only $100. It is made of clear a1/4" thick acrylic, about 3 feet wide, 3'
high and 2' deep (18 cu ft or 2 cu meters) with a vacuum airlock on the side and two nice long rubbery gloves in the front. It also came equipped with
hose connections and flow meters. Such a deal!
The idea was that I could fill this thing with N or Ar and then do all those experiments I have been wanting to do but can't: reductions with LiAlH4,
or playing with alkali metals, or such. I figure the box can also double as a fume hood for some nasty reactions in that the inside atmosphere is
totally isolated and can be vented somewhere far away, like the neighbors yard ;-) (No. really it is a hard question. Where would I vent it?)
The problem and subject of this post is: what is a good way to make sure all the oxygen has been swept from the chamber? There is an inlet port and an
outlet, but you can't, say, remove all the air and then pump something new in because the chamber would collapse.
A few different methods came to mind:
1. Theoretical: If I pump in N2 at a certain rate, x liters per minute, and vent from the chamber at the same rate, and assuming good mixing of the
incoming N2, then the O2 level inside the chamber should fall at an exponentially decaying rate. This is because the rate of decrease in partial
pressure of O2 is proportional to the partial pressure of O2, and that is the hallmark of exponential decay. So I could just compute the amount of
time it would take have faith in my computations.
2. Burn a candle in the chamber (or fan a glowing ember or whatever) until the candle burns no more. Other, less smoky equivalent would be to make up
an O2 adsorbing solution such as pyrogallol in alkali.
3. Any combination of the above plus, put some kind of O2 sensor either inside the chamber or in the exhaust stream from the chamber (assuming N2 is
being actively admitted at some rate most the time.)
Problems with above:
1. Because the fall of partial pressure of 02 is exponential, it can take a long time to reach a very low level. And waste a lot of N2just to get from
1% to 0.1%. It is better to have a more proactive way of reducing the O2 to nearly zero once the level gets low, say below 1% of atm. That is exactly
what method 2 is good at, but….
2. A candle is kind of a cute idea,but when it goes out it makes a lot of smoke and I am not sure I want that contamination. An ember would not
produce smoke but like a candle, but would produce C02. That would not be suitable for some chemical reactions. C02 will happily combine with many
strong reducing agents. So that leaves the pyrogallol in alkali solution (or other similar solutions). These are good for adsorbing a relatively small
amount of oxygen, and one can compute how much solution wood be required after flushing with N2 for so many minutes, BUT, in my experience once you
have that many variables real world unknowns intrude and I wouldn't really trust 100% that everything had gone well . So that is what #3 is for…
3. So a quick reading cheap oxygen sensor would be really nice for positive verification that the O2 level is sufficiently reduced to few enough
milli-moles to not interfere with the reaction at hand. But how to make one?
The first thing I tired was a solution of ferrous sulfate and potassium thiocyanate. I read that ferrous ion is oxidized to ferric ion in air. And
ferric thiocyanate is deep red. Since I had these chemicals lying around I mixed up the solution to see what it would do.
I added about 15 g FeSO4 to 100 ml H20 and then added 1g KSCN.
When first mixed the solution is light green as expected.
As a test, I added a drop of 3% H2O2 to a sample and it immediately turned dark red. Good.
On exposure to air the solution gradually turned orange over the course of an hour. And overnight it did , in fact, turn dark red.
Well that's a nice chemistry experiment, but much to slow for indicator. BUT the bigger problem, the one I did not fully anticipate is that the
reaction needs to be reversible if the solution is to be a good indicator. If I then bubble pure N2 through the solution, ideally it would turn back
to light green. Intuitively, I did not think this would happen. But I tried the experiment anyway. And, no, it did not happen. Stayed red.
So. I tried to think of, or research, a reversible reaction with dissolved O2 that could be detected either via a color indicator or
electrochemically. I had a brain storm. Blood! Blood turns red when oxygenated and for hemoglobin to be useful it has to give up the O2 when the
partial pressure falls below a certain level (and then turns some other color). Can one buy hemoglobin? No. Not really, not for cheap. And I suppose I
could use a small sample of my own blood, but I would have to keep it from coagulating. Anyway the whole thing is sounding a bit messy and ghoulish.
(But hey, Halloween is coming up.)
I had second brainstorm: fuel cells! I could make a tiny fuel cell that depended on O2 to produce current. Further research pinpointed a most useful
fuel cell which is really a battery: a zinc air hearing aid battery. These will only produce a current when exposed to O2. By choosing an appropriate
load resistor the voltage will go to 0 as the oxygen partial pressure falls below a specified amount. I am not sure what the response time will be.
The holes on the negative terminal of the battery are very tiny. But, it is much less messy than a vial of blood with gas bubbling through it. So I
bought 8 tiny zinc air button batteries for $9.
The only problem was I needed a good battery holder so I can check the current while the battery sits inside the glove box. So I haven't been able to
test the batteries yet. I went to the DigiKey website to order some battery holders and they haven't arrived yet. The ironic thing is that while
browsing through the DigiKey online catalog I noticed that they have a section on sensors, and in that section there is a sub section on electronic
gas sensors. And low and behold in that subsection they had O2 sensors. BUT these O2 sensors cost from $300 up, so I still feel vindicated that my
zinc air O2 detector will be a useful idea. But I might buy a CO2 detector from DigiKey. They are only about $20. CO sensors even cheaper.
Anyway. If any of you have good ideas on O2 sensors I would like to hear about them.
(Note: I did notice that one can buy O2 sensors relatively cheap (~$40) which are used by car computers for sensing oxygen in automobile exhaust
gases. But I could not find enough information about how they work or how to use them so I was reluctant, but keeping this option in reserve. Also
they operate at very temperatures (300C) and that is a drawback.)
[Edited on 28-10-2011 by wxyz]fledarmus - 28-10-2011 at 10:52
You may not be able to pull a really hard vacuum on the box, but purge-and-fill is still a very useful way of replacing atmospheres. It may cost you a
bit in N2, but not nearly as much as your constant venting method number 1.
The idea is to pull a low vacuum on your box to remove, say, 90% of the air (down to 1.4 psi). You then allow N2 to fill the box back to atmospheric
pressure, and pull a low vacuum again to remove 90% of the air. You have now removed 99% of your original oxygen. Fill and purge, and you have removed
99.9% of your original oxygen.
If your box won't hold at 1.4 psi internal pressure, it will simply require more cycles. Knowing how low a vacuum you can hold, you should be able to
calculate how much N2 it would take to purge to whatever level of oxygen you feel comfortable with.wxyz - 28-10-2011 at 18:38
Thanks for the reply. I wasn't so much worried about pulling a hard vacuum as the box collapsing under the weight of the outside air. At atmospheric
pressure of 15 psi, or 2160 pounds per square foot on the outside and 1/10 that on the inside, I am pretty sure the thing would collapse. Even 1/2 an
atmosphere inside would probably not be enough to keep it from imploding.unionised - 29-10-2011 at 05:51
You can't suck a useful amount of air out of the box to purge it. But there's another useful trick. Get a big trash bag and put it in the box.
Feed a pipe into the bag so you can inflate and deflate it. (you will need to cut a hole in the box to do this).
Drill another hole in the box and connect a pipe to it.
Close the box and inflate the bag.
The air in the box will be forced out through the second pipe by the bag. If the bag is big enough it will take up all the space in the box.
Once the bag is full, connect your nitrogen supply to the other pipe. The box will be filled with nitrogen. (and the air in the bag will get driven
out through the pipe).
That way you can fill the box with nitrogen using just one "boxful" rather than several "boxfuls". antibody - 29-10-2011 at 12:55
have you considered using argon, its much heavier than air, so it will fill your chamber from the bottom up, displacing air, (i think it will mix less
than nitrogen) so if your air outlet is at the top this might work too. I like the garbage bag idea too.
[Edited on 29-10-2011 by antibody]benzylchloride1 - 29-10-2011 at 22:21
This reference describes the preparation and use of a titanocene complex for use as an oxygen indicator, this may work if you can obtain titanocene
dichloride. Journal of Chemical Education • Vol. 75 No. 4 April 1998. A good way of flushing the air out of a glove box is by inflating a trash bag
with nitrogen inside of the glove box and forcing the nitrogen out by collapsing the trash bag inside of the glove box several times.unionised - 30-10-2011 at 01:51
This reference describes the preparation and use of a titanocene complex for use as an oxygen indicator, this may work if you can obtain titanocene
dichloride. Journal of Chemical Education • Vol. 75 No. 4 April 1998. A good way of flushing the air out of a glove box is by inflating a trash bag
with nitrogen inside of the glove box and forcing the nitrogen out by collapsing the trash bag inside of the glove box several times.
That's a fairly good way. A better way is to use air (which is cheap) to inflate the bag. Also, if you use a big enough bag then you only need to do
it once, so you save even more gas.
I'm sure I saw that described somewhere recently.Harristotle - 30-10-2011 at 02:04
Hi.
My first post, so please be gentle!
My favourite oxygen sensor is a hearing aid battery (alkali zinc) with a 1k resistor across it. Measure the voltage drop across the resistor.
Current (of battery) is proportional to concentration of reactants, and O2 is limiting so therefore current is set by oxygen concentration. The
voltage drop across the resistor is set by the current flowing through it, via Ohms law.
I haven't tried this myself, but I reckon that you could modify a 10ml syringe to calibrate it with (1, 0.5, 0.25 x atm pressure).
Finally, I wish to acknowledge this forum. I have lurked for a long while, and the high quality of the discussions, together with their practical, OTC
focus, has greatly informed and helped my high school teaching practice. Keep up the good work!
H.wxyz - 30-10-2011 at 09:48
Thanks for all the replies!
The garbage bag method sounds great: simple and efficient. I am going to give that a try.
Also good to have the the titanocene dichloride reference.
Although it is not clear if the reaction is reversible, at least it is a lot faster than the one I dreamed up (FeSO4 + KSCN).
The Ar idea is interesting. It may work, and in fact my box does have one port near the top. I imagine one would have to let the Ar in slowly to be
sure that there wasn't too much mixing. And in this case a sensor would be essential as it would be very hard to predict how much mixing took place
and if the air was really totally expelled.
Lastly, thanks Harristotle for confirming that Zn-air batteries can be used as O2 sensors. I had suspected as much (its buried in my original post)
but hadn't collected all the parts to test it yet. matei - 30-10-2011 at 10:47
Hi.
My favourite oxygen sensor is a hearing aid battery (alkali zinc) with a 1k resistor across it. Measure the voltage drop across the resistor.
H.
This is an interesting idea, however it has a limitation: the battery uses an aqueous electrolyte, so in a very dry atmosphere such as in a glovebox,
the water will evaporate quickly and the current will drop accordingly.Harristotle - 31-10-2011 at 03:38
Quote:
This is an interesting idea, however it has a limitation: the battery uses an aqueous electrolyte, so in a very dry atmosphere such as in a glovebox,
the water will evaporate quickly and the current will drop accordingly.
Could this be improved by adding a drop of glycerol to the battery - it is hard to evaporate glycerol/water mixtures and these are used to prevent all
kinds of things from drying out?matei - 31-10-2011 at 09:25
IMO a Glycerol/water electrolyte would be inefficient because it has a high viscosity so ion mobility will decrease - think of the car battery which
doesen't work properly at low temperatures for the same reason (more viscous electrolyte).
Anyway, although the principle of using such a battery as oxygen sensor is OK, I don't know if it is suitable for measuring such low concentrations of
O2 (< 0.001% or whatever) because in this case the current would be too low and you'll have to use a special amplifier circuit (maybe a very
sensitive electrometer).unionised - 31-10-2011 at 11:47
I haven't checked but, to be much use as a power source the battery would need to be able to produce a few mA in air with 20% oxygen. (my guess is
tens or even hundreds of mA)
I think the current is proportional to the concn.
So if we drop the concentration to 1ppm the current will fall to at least few tens of nA and maybe even as much as a microamp.
That's not difficult to measure.
The drying out problem is an issue, but those batteries come with a little sticky tab on them that keeps them sealed when not in use so you could keep
them closed up most of the time. That would limit the drying out.ziqquratu - 31-10-2011 at 22:35
Simplest oxygen detector I've seen in a glove box is an incandescent light bulb. Heat the bulb carefully over a flame to create a hole in the glass
(be ready for a loud pop!), then put it in the box. If there's a notable level of oxygen in there, the now-exposed filament will smoke (and eventually
burn out completely - or do so almost instantly if the level is high). Obviously not quantitative, but useful nonetheless - and extremely cheap and
easy!
As for purging the box - I believe the garbage bag idea is the best way to do the initial purge. For maintaining the atmosphere, we use a constant
small positive pressure of inert gas, with only a slow flow rate. Inside the box is a little recirculating pump (aquarium pump or similar would
probably get it done), which passes the gas in the box through tubes full of activated molecular sieves (to remove water) and an oxygen scrubber (I
believe a DIY version could be made using chunks of activated charcoal impregnated with copper - I'm pretty sure I've seen a method online).wxyz - 2-11-2011 at 21:34
I haven't checked but, to be much use as a power source the battery would need to be able to produce a few mA in air with 20% oxygen. (my guess is
tens or even hundreds of mA)
I think the current is proportional to the concn.
So if we drop the concentration to 1ppm the current will fall to at least few tens of nA and maybe even as much as a microamp.
That's not difficult to measure.
The drying out problem is an issue, but those batteries come with a little sticky tab on them that keeps them sealed when not in use so you could keep
them closed up most of the time. That would limit the drying out.
I totally agree, even 1 nA is not all difficult to measure using a $1 op amp. And these days with a slightly more expensive op amp ($3) you can
measure 1/100 of that accurately. Not sure if the linear relationship between O2 conc and current holds down the far, but would be interesting to see.
As far as drying out is concerned, these things have to work in a dry place like Arizona, where I am sure there are a lot of hearing aides :-) So
that, along with putting the tape back, would seem to be adequate.
