Ice in a Glass-Lined Vessel

Find a glass receptacle that is dispensable, place ice in the receptacle, and agitate it for a prolonged period of time.

If the glass is undamaged, there won't be a problem.

If the glass is damaged, you've only ruined a throw away container.

Ice is not hard enough to scratch glass - unless there are carried hard particles to scratch the glass. If the glass is thick and there is carried water and other reactants, you should be OK as long as the vessel will bear the mass impacting.

The glass will not abrade but the pounding of the ice will cause the vessel to flex, possibly dent, and that could cause the glass to delaminate and/or break.

How about working around the potential problem and using soft ice?

Seriously, can you use shaved or crushed ice?

-------------------------------------------

be/hatz/la/cha

The problems are

1. Large blocks of ice might damage the agitator, unless made for the task.

2. melting of the ice will dilute the reagents

3. I assume you are catering to a James Bond look-a-like convention and need a large volume of shaken-not-stirred vodka martinis?

4. Perhaps you will advise the true application? Diazo reactions?

I am quenching quite an exothermic reaction. A Friedel Craft Reaction. I have utilised my utility as much as it can provide in terms of flow rate and temperature. My quenching is still very slow. My limitations are quenching in water (not methanol) and the other solvent is Dichloromethane so I can't use a Polypropylene agitated vessel with ice (usually used for such quenchings).

We are thinking of crushed ice, yes, but I don't want my glass lining to chip.

I wish it were a martini!

:)

Makes me recall my organic labs where we messed with this stuff and the profs watched like hawks to make sure no speed freaks were in the classes - LOL.

I think you should have recourse to the organic chemistry library of a major university, and find people doing similar work to yours and see how their thesis procedure sovled their quenching problem.

Some of this data will be online, behind a paywall. If you are a college student your student ID will get you past the paywall if you work in the library network.

I just gave Miklosh (#4) a GA because he was going where I was.

Crushed Ice will create a slurry which will inherently ALMOST assure that you will have liquid between the ice and the vessel wall, plus provide a massive (relatively) amount of surface area for the heat transfer which will provide quick cooling and even softer ice before it melts completely.

Crushed ice is also much easier to control the temperature because you can easily calculate the mass of ice needed to cool your process solution by just dividing the thermal cooling load by the combined latent heat and sensible heat from the ice melting point to the target temperature. As noted earlier, the high surface area will trigger a fairly quick cool down- much faster (and safer) than large blocks or chunks of ice.

The final advantage to crushed ice is the ability to "inject" it simultaneously with the solution you are cooling if this process could benefit from "continuous flow" vs. a batch process.

Good luck. By the way, crushed ice is also the fastest way to get the "shaken, not stirred" dilution while cooling effect that worked so well for Mr. Bond.

And the glass used in lining vessels is different from laboratory glassware. Usually lower Silicon oxides here.

did you try searching on the website for the manufacturer of the agitator ?

they might list upper and lower temperature limits

are you asking here because you need your answer before starting work in the morning or did you already try the aforementioned option ?

some materials have different resistance to impact depending on their temperature , plus ice might cause excessive shrinkage and delamination

if the constituents are not mixed evenly and you got hot and cold zones in the agitator that could cause cracking

Why glass-lined? Is there another material from which the vessel may be made without affecting the reaction?

How large is the vessel? Would converting the process to a continuous stream be better in terms of heat removal opportunities?

Is there a Process Engineer involved with the development? What can this individual bring to the party?

I am the process engineer in question.

The reactor is a Mild Steel Glass Lined Batch Reactor, top agitated. It has a working volume of 5kL. Nope, there are different grades of glass lining, but their resistance to impact by ice is not the best publicised. The manufacturer suggests ice shouldn't be used- crushed or large pieces. We are limited by the heat transfer coefficient of this vessel, the jacket has been descaled (unfortunately hollow hence lesser utility velocity, lesser Heat transfer), we have more then adequate utility pressure drop across the jacket though. The utility temperature is limited by the refrigeration unit capacity. We have a retreat blade turbine as the agitator that is amongst the best radial-flow turbines and provides the highest wall velocity for excellent heat transfer.

Converting to a continuous mode would be the best solution, but continuous manufacturing is not the most common episode in Pharmaceutical Manufacturing. Convincing organisations to shift to Continuous Manufacturing and getting them through the Regulatory approval for this is a part of my daily job. But oh, the travails of Pharmaceuticals.

I guess we are going to make do with a slow quenching operation, unless we add another reactor.

Thanks for the detailed summary of your operation zasmine

What is the cooling medium (you called it utility?) that you pump through the vessel's jacket? Temperature in/out?

What you need to do as use an ice shaver within the vessel or through a port where the ice is shaved/chipped to 1 - 2 cm size. This will minimize or eliminate wall damage. What is the rate of heat production? The chipped ice input must exceed this, or thermal runaway may occur = spolied batch.

The easiest way to solve your Heat-Ex issue is to use a primary-secondary pumping system for the cooling fluid. Primary is the current feed pump and secondary would be a new pump connected to the inlet and outlet of the jacket cooling solution connections.

You should be able to calculate the flow required to produce significant turbulence.

For example- if your current solution flow is 10, with a jacket entering temperature of 7C and an 8C DeltaT (leaving fluid temp of 15C), and the calculated "turbulent" flow is 80- then your new system would circulate 80 through the jacket and the jacket's DeltaT would now be 1C. The "down-side" of this system is that the temperature of the fluid entering the jacket can be no colder than 14C so that the existing chiller will still see 15C entering (and the 7C leaving temp would mix with the 70 units of 15C recirculated fluid to create 80 units of 14C entering fluid- not necessarily the best from a cooling mode, unless the batch being cooled has a final temperature well above 14C.

Since you were talking about possibly using ice, you would have had to install added refrigeration capacity any way. If you use that new refrigeration to make a very dilute brine solution (whatever product works for your application) at, say 1C leaving temperature (or whatever temperature is needed to avoid any possibility of freezing the materials in the batch being cooled) and you circulate the new 80 unit flow through that new "process chiller", you could use your existing 10 unit flow from the chiller as the cooling medium for the process chiller. It will see a 9C DeltaT but the added heat should not mess up the overall system unless that existing chiller is dedicated to this process.

Should THAT be the case, by raising the leaving water temperature from 7C to, say 12C (21C return fluid) will definitely increase its cooling capacity enough to support the new load.

By the way, although the new brine would not be below freezing, the use of the dilute brine is a cushion to eliminate any risk of that solution freezing due to out of calibration controls or accidental reduced flow without paying a huge penalty for reduced thermal capacity or increased viscosity.

No experience here so sorry if this is a stupid suggestion.

Would using LN₂ (liquid nitrogen) be over the top/too uncontrollable for this application. It's cheap and readily available.