Sciencemadness Discussion Board

Glass adapter pieces to put a flask under pressure?

JefferyH - 30-5-2014 at 12:47

I have a reaction that I want to put under pressure with Nitrogen so that it can be heated to a higher temperature without refluxing to speed the reaction. I know the borosilicate glass can handle up to 100 PSI.

My concerns are what sort of adapters I will need to do this.... Removing pressure is one thing since the glass is being pulled towards the vacuum, but putting something under pressure, all of the adapters will be pushed outwards. Are normal plastic clips strong enough to hold onto the adapters with these pressures? What about a gas inlet adapter? Do I need something that has a one-way check-valve mechanism?

I only plan to go up to 40 PSI roughly, since I just need to bring the water to around 130 degrees without it refluxing. How significant is this amount of pressure, and what glass pieces are advised for this?

jock88 - 30-5-2014 at 13:09


If you can spare some glass and you have spare glass you could set up with just water at 130C or perhaps 150C and put pressure to 100psi, stand back and see what happens.
The statement 'borosilicate will stand up to 100psi' is misleading as it depends on the thickness of glass and size/dimensions of glass vessels.

Muffn Man - 30-5-2014 at 13:53

I know you said glass, but I don't know of any glassware that will withstand pressure as opposed to vacuum. The way most labware is designed, you would need to seal up all of the joints with electrical and/or duct tape for it to even stay together.

Is a pressure cooker an option for you? They are made to be pressurized with inter-locking parts and proper seals. Just a thought.

[Edited on 30-5-2014 by Muffn Man]

JefferyH - 30-5-2014 at 13:58

Quote: Originally posted by jock88  

If you can spare some glass and you have spare glass you could set up with just water at 130C or perhaps 150C and put pressure to 100psi, stand back and see what happens.
The statement 'borosilicate will stand up to 100psi' is misleading as it depends on the thickness of glass and size/dimensions of glass vessels.


I understand the glass thickness is a factor, the glass I am using is very thick walled. As for the first part of your post, I have absolutely no clue what you said..... water cannot exceed 100 degrees unless under pressure first.
Quote: Originally posted by Muffn Man  
I know you said glass, but I don't know of any glassware that will withstand pressure as opposed to vacuum. The way most labware is designed, you would need to seal up all of the joints with electrical and/or duct tape for it to even stay together.

Is a pressure cooker an option for you? They are made to be pressurized with inter-locking parts and proper seals. Just a thought.

[Edited on 30-5-2014 by Muffn Man]

I thought about a pressure cooker but I don't think I can stir through one. I'll have to look into that.

Most glassware that can withstand a vacuum can also withstand being put under pressure. Lab grade borrosilicate glass is generally considered to withstand 100 PSI. I think even plastic soda bottles can withstand that much.

unionised - 31-5-2014 at 03:03

" Lab grade borrosilicate glass is generally considered to withstand 100 PSI. "
No.
It is not.
Do not post dangerous nonsense like that.

Muffn Man - 31-5-2014 at 09:19

I think that glassware is better at vacuum than pressure. The rounded shape acts much like an arch made of bricks. The arch can hold up under a lot of weight/gravity but it would not work if it were inverted (ie pressure in a vessel).

An aluminum pressure cooker would allow you to use a stir bar and a magnetic stirrer. They can be had for very little money at thrift stores or garage sales.

unionised - 31-5-2014 at 11:32

Pressure cookers are typically rated for 15 PSI, not 40.

Muffn Man - 31-5-2014 at 12:10

"Pressure cookers are typically rated for 15 PSI, not 40."

Pressure cookers are regulated to 15 psi.

If somebody were to disable the regulator or modify it to hold more pressure, the rating, as you say, would go up.

15 psi is not necessarily the maximum that the vessel will handle.

The manufacturer will only accept liability for any harm caused by the device when it is used in compliance with their guidelines. When the device is used according to the guidelines, it is quite safe.

Rogeryermaw - 31-5-2014 at 12:25

if your reactants don't react with steel or zinc, it would be relatively cheap and easy to build a suitable reactor for pressurization from some rigid pipe, fittings, a cap and a valve. you could also use a tee fitting if you need two valves if, for example, you needed to have a gas line hooked up and a port through which you could add reactants or remove product.

unionised - 31-5-2014 at 13:06

Quote: Originally posted by Muffn Man  
"Pressure cookers are typically rated for 15 PSI, not 40."

Pressure cookers are regulated to 15 psi.

If somebody were to disable the regulator or modify it to hold more pressure, the rating, as you say, would go up.


No, they are rated for use at 15 PSI, and they will be, no matter what you do to the valve.
They will, in fact, stand more than that but the point remains that they are not designed for use above 15 PSI.

smaerd - 31-5-2014 at 13:39

I think rogeryermaw has the best idea here. Always over-spec apparatii like this. Meaning if you calculate you need a quarter inch of some grade of stainless steel, make it a half inch. So many unanticipated thermodyanmic things happen with chemical reactions, even very well studied ones. If you calculate you need a liter of volume above some liquid reactants. Make it 1.5L or more... Always incorporate fail-safes, blow off valves, etc.

If you're simply pressurizing water, then yea you should be fine with steel or zinc as roger has said. I'm not sure that I understand the goal exactly. Could you elaborate a bit more on what you are trying to do/why? If you plan to flow this water somewhere through an out-let there will be a pressure decrease depending on how it is performed... Also not sure what you mean by "how significant is this pressure"? It's a relatively high pressure, same as being about 95 feet under water, about 2.75 times the total pressure of the earths atmosphere(at sea level, mean value, whatever).

Not that I am a pressure/engineering expert at all but yea, don't go above a pressure cookers specifications. Don't apply 100 PSI to laboratory glass-ware either... 40PSI sounds like too much for glass but it's not like I've checked it. Solid 10PSI greater then what people put in their car tires, any mechanical failure at that pressure could be hazardous of course. I have seen in old-school synthetic preparations where people were doing reactions under significantly smaller pressures in their glass-ware. The joints were wrapped shut with some kind of wire I believe. Actually the example I am thinking of the glass-ware was stoppered and wired shut, no other joint-ware.

40PSI with high temperature I'm assuming? Most plastic clips are not going to last out. Also how much water are you trying to get to this pressure/temperature?

[Edited on 31-5-2014 by smaerd]

Dr.Bob - 2-6-2014 at 11:40

Normal glass labware is not designed to be used for pressure, especially jointed glassware, as the joints are designed to hold vacuum, not pressure. There are high pressure glass tubes for chemistry that can handle 100-300 psi, when used properly, but those are designed to be sealed at normal pressure and temp, and then heated to above the solvent's BP to create the pressure. This should be done behind a shield and using some way to monitor and regulate the pressure. Chemical microwave reactors are a good example of that type of system, which uses IR sensors and pressure transducers to allow small reactions to be heated to higher temps and pressures in special glass tubes.

The only routine reactions that have external pressure applied are hydrogenations, which require special glassware and equipment, and are normally done at up to 40-60 psi in glass, over that is normally done in special equipment. Most other laboratory chemistry with gases is done at mild pressures, except for specialized labs; even CO insertions are rarely done at higher pressures.

Without knowing the reaction, it is hard to suggest useful ways to do it, but most reactions can be done at lower pressures if the solvent is changed to a higher boiling temp, the reaction is run longer, or a catalyst is used to make it work better at routine pressures. Industry does many reactions at pressure, but rarely in glass flasks.

Praxichys - 3-6-2014 at 17:49

Remember that pressure is force over area... a 10-inch diameter pressure cooker has a lid surface area of about 78 square inches, and will thus be pulling on the rim with 78 pounds force for every PSI of pressure in the cooker.

@ 15 PSI(g) the lid is trying to come off with 1170 pounds force
@ 40 PSI(g) it is over 3100 pounds force

Would you feel safe hanging a small car from the lid of a pressure cooker full of hot liquid in a small room?

JefferyH - 3-6-2014 at 18:46

Good points. I probably won't be trying it, I just wanted to do it as an experiment. Nothing like a hydrogenation or anything like that. I wanted to run several different types of reactions that occur at known speeds at known temperatures, and see how much the speed will increase if the temperature is increased by 30 C or so.

Using higher boiling point solvents are one option, but seeing as I need to dissolve large quantities of salts for these simple reactions, DMSO or similar solvents would be too costly in the quantities needed, compared to the amounts of water needed. I've always disliked the use of DMSO for a number of reasons, these included.

Not really looking to invest in a pressure vessel right now. The inability to stir would defeat the purpose of the experiment. A pressure vessel with built in stirring would cost way too much.

Oh well. Not a big deal. Maybe I'll do this later when I have better equipment.

Praxichys - 4-6-2014 at 04:16

You could try building something like a shaker hydrogenator. A steel vessel wrapped in a heating mantle rests on a pivot point and a motor oscillates the vessel back and forth to shake the contents. Pressure, heat, and feedback systems for pressure and temperature are easily connected via flexible cables and tubes. Ordinary 2" schedule 40 steel pipe from the hardware store (Biggest I can get locally) has a working pressure of 1500 PSI and a burst pressure of ~7800 PSI.

Just be careful. Remember that reaction rates increase with pressure and temperature. It is easy to turn a reactor into a pipe bomb by accident. I would reccommend a built-in pressure relief valve or burst disk, safety shields, and a "just-in-case" plan.

For what you're doing, however, I'd imagnine you would like a way to observe the contents?

Dr.Bob - 4-6-2014 at 05:13

You can get small glass pressure tubes easily and affordable. They would allow you to stir with a magnetic stir bar. But most of those are only for small scale, typically 5 ml to 30 ml are common, there are some larger ones, but they cost more. But those work well in an oil bath and allow about 10-30 C increase in temp for water.

JefferyH - 4-6-2014 at 09:19

Quote: Originally posted by Dr.Bob  
You can get small glass pressure tubes easily and affordable. They would allow you to stir with a magnetic stir bar. But most of those are only for small scale, typically 5 ml to 30 ml are common, there are some larger ones, but they cost more. But those work well in an oil bath and allow about 10-30 C increase in temp for water.


That is of course one solution but this will severely effect my ability to measure the reactants, and I do not have a professional lab equipped to measure super small samples, just a home lab.

Smaller reaction sizes would result in:
- Greater difficulty in measuring products vs substrate ratios
- Requirement for a smaller heat source, without a heating mantle that I would be using, unreliable/inconsistent heat input might result.
- The physical shape of the tube even with stirring would altar the substrate distribution/mixing - this would also have an effect on the reaction.

There are other things, though, the primary thing is due to my lack of super precise and super expensive chemical analysis equipment, and lack of small-scale reactors, the results of a small reaction would differ too greatly in too many potential ways. Maybe I am over exaggerating how much the reaction would be influenced, but all these factors are important. For my home lab setting, even with small glassware, I typically do not like to run reactions with less than 1/4 mol quantities. For reactions that aren't messy, very small vessels under 100ml could work, but any bigger there is the concerns that other people have been sharing about glass outward breakage from pressure.

unionised - 4-6-2014 at 11:10

"Maybe I am over exaggerating how much the reaction would be influenced, but all these factors are important."
Yes, and also you missed an important factor
"Smaller reaction sizes would result in:"
Less stored energy when something goes wrong.

I don't know what your lab facilities are like but this sort of thing worked just fine over a hundred years ago.
http://en.wikipedia.org/wiki/Carius_halogen_method
and they didn't have the sort of kit people rely on today.

It's obviously obsolete now, but it shows that one can do the job perfectly well if you know what you are doing.

However, when it comes down to it, if you are so ill informed as to say "Most glassware that can withstand a vacuum can also withstand being put under pressure. Lab grade borrosilicate glass is generally considered to withstand 100 PSI." you probably shouldn't be trying this sort of thing at all.



Dr.Bob - 4-6-2014 at 12:24

I use small reactions all of the time with a simple heating mantle or oil bath and TLC. Not everyone works in the highest tech labs nowadays. The basic guideline I learned is that reactions go twice as fast for every 10C degrees of higher temp. That roughly holds true for the ones that I have tested. So if you run the reaction a set time at a set temp/pressure, you should be able to compare that with the same reaction at a different temp or time, whether it is 10 mg or 10 grams. And you can use these tubes with a heating mantle, and a sand bath, done it before myself, but oil bathes are fine. But the best way is a special microwave reactor, they rock for small scale pressure reactions, the best use of microwaves since potatoes...

As for stirring, if the reaction is heterogenous, then stirring may be as or more critical than temp, and you are right that stirring is critical. That is where the Parr type shaker is good, or a paddle stirrer, which is REALLY hard to use under pressure, without special equipment. If the reaction is homogenous at the set temperature, then stirring should not have a huge effect on rate.

The experiment you describe sounds interesting in theory, but hard to judge how practical without more details. And while some glassware might be OK for 100 psi in some cases, I would only try that in a very carefully set up, shielded situation, and only on small amounts. Large scale metal pressure reactors are often called "bomb" reactors for a good reason. I have used a few and they are a pain to use and clean, but do have blow out ports to prevent the worst case scenario. Exotherms in those can be very, very bad.

JefferyH - 4-6-2014 at 12:36

That is precisely what I am testing, the idea that reactions go twice as fast for every 10 degrees. I am skeptical that this rule holds true for all reactions, or at least, has variation depending on the type of reaction. Part of the reason I want to test it is to help gain more confidence in the idea, instead of just assuming that raising the temperature will increase the speed of my reaction.

I could always do a low-dollar rig.... glass champagne bottles are designed to hold 100+ PSI, but not much more. I could outfit one with a pressure relief valve and set it to automatic relieve if the pressure exceeds a certain something. This combined with a hotplate or torch and a paint shaker. Lol, but I can't see this getting results that are very close to a legitimate set up. Maybe for a ghetto hydrogenation vessel, but not much more.

Lambda-Eyde - 4-6-2014 at 12:59

Quote: Originally posted by JefferyH  
I could always do a low-dollar rig.... glass champagne bottles are designed to hold 100+ PSI, but not much more. I could outfit one with a pressure relief valve and set it to automatic relieve if the pressure exceeds a certain something.

You don't need a pressure relief valve, soda-lime vessels are perfectly capable of relieving themselves of pressure when heated.

NEVER use anything but genuine heat-resistant lab glassware for anything involving heat. From what you write, I can tell that you're way too inexperienced to be experimenting with reactions under pressure. Please stop before you hurt yourself - what you're proposing is essentially making a pipe bomb with a chemical payload. Do something safer, gain experience, read a book (or twenty) and learn routine procedures before advancing to more demanding (and dangerous) procedures.

phlogiston - 4-6-2014 at 14:32

If I understand correctly, your test reactions take place in water.
Then, the range of (approx) 0-100 deg C is experimentally accessible without special effort.
It seems not very likely to me that the additional 20-30 deg C extension on the high end of the scale is going to matter much for your conclusions regarding the effect of temperature on reaction rate, and then perhaps it does not warrant the extra effort.

If you are desiring to test the general relationship between reaction rates and temperature and see possible deviations from the trend at extreme temperatures, perhaps it would be easier to find a reaction that takes place in a different solvent.

JefferyH - 4-6-2014 at 19:33

Quote: Originally posted by phlogiston  
If I understand correctly, your test reactions take place in water.
Then, the range of (approx) 0-100 deg C is experimentally accessible without special effort.
It seems not very likely to me that the additional 20-30 deg C extension on the high end of the scale is going to matter much for your conclusions regarding the effect of temperature on reaction rate, and then perhaps it does not warrant the extra effort.

If you are desiring to test the general relationship between reaction rates and temperature and see possible deviations from the trend at extreme temperatures, perhaps it would be easier to find a reaction that takes place in a different solvent.


I have a book in my car with some test reactions I wanted to. I recall one reaction takes 6 hours to finish at reflux (of water). If raising that temperature a mere 40 degrees would bring the reaction down to 22 minutes... then that is something I would be interested in seeing first hand. An extra 40 degrees is nothing for a decent heating mantle, and if the relationship between temperature/reaction rates is even halfway true, then that's just simply amazing. I almost don't believe it.

unionised - 5-6-2014 at 11:27

Quote: Originally posted by JefferyH  
That is precisely what I am testing, the idea that reactions go twice as fast for every 10 degrees. I am skeptical that this rule holds true for all reactions, or at least, has variation depending on the type of reaction.


Well, that's easily solved.
The "double the rate for every 10 degrees" rule is certainly not valid for all reactions.
If you look here
http://en.wikipedia.org/wiki/Arrhenius_equation
you can see why it's roughly true- there's a lot of reactions with similar activation energies.

So, there's no need to take the risk of doing all the experiments.

Incidentally, champagne bottles are not designed to be heated so don't use them at anything other than room temperature.