HPLC leak at tubing/column interface

[ZeroAir]

Hello,

I just installed a used C18 column on our lab's HPLC, and was pumping 95:5 Water/ACN when the pressure rose and the connection between the tubing from the autosampler and the front of the column popped out. I tried to re-tighten the ferrule, but it still pops out when I resume pumping.

I have tried new ferrules, 100% ACN, and have thoroughly dried the interface between the ferrule and column. I have also tried roughing up the interface with sandpaper in accordance with a suggestion on another thread of this forum.

Does anyone have any suggestions?

Sincerely,

Luke

[tom jupille]

First question:
What kind of tubing and fittings are you using (stainless steel or PEEK)?

Second question:
How high did the pressure go when the the tubing popped off?

Third question:
What was the pressure when you resume pumping?

[ZeroAir]

Fittings are PEEK, at ~150 bar the tubing pops off, I can replace the fitting and start the pump pressure goes right back to 150 and pop!

Thanks for your help,

Luke

[Markus Laeubli, Metrohm]

Try to roughen the capillary surface with e.g. a abrasive paper, or similar.
Another possibility would be to use a tool even if the fittings are called 'fingerthight'.

[tom jupille]

PEEK tubing and fittings should hold to at least twice that pressure. However, pressure gauges have a lag time and if your system rapidly pressurizes, the actual pressure may be significantly higher than that, which raises the possibility of a plugged flow path.

Definitely follow Markus Laeubli's advice about the pliers (you have already tried the sandpaper trick), but then restart the system at a lower flow rate and then gradually increase it while monitoring the pressure.

[ZeroAir]

Alright, slowly increasing pump rate did the trick!

Here is a plot of pump pressure vs time as I increase the flow rate. Max pressure was 165bar at flow rate 3.0mL/min.


free image hosting

Is it possible that the initial decrease in pressure to steady state indicates a clearing in whatever blockage may have been present?

Thanks again.

Luke

[tom jupille]

Is it possible that the initial decrease in pressure to steady state indicates a clearing in whatever blockage may have been present? Thanks again.

Yes it is. Normally you would expect the initial step to look like the others. Now the trick is to make sure you filter samples and mobile phase well enough to prevent recurrence!

[ZeroAir]

Thanks again.

L

[lmh]

there is another explanation for the first step looking different. If the column has been stored in an organic solvent, and you start to purge it with a highly aqueous phase, you are gradually changing the solvent in the column. Different solvent mixes have different viscosities. 50% methanol, for example, is more viscous than either water or 100% methanol, so a column stored in methanol, when flushed with water, will give a gradual rise in pressure and then a decline not unlike what you showed.

I am deeply dubious about sandpapering peek tubing. An as-new finger-tight Peek fitting on as-new tubing should hold much more than 150 bar unless you suffer from weak fingers or it's in a particularly difficult place to tighten. Whatever happens you don't want to run any risk of getting powderised Peek into the top of your column; it will get stuck in the frit and will give you high back-pressure for ever.

I'd also urge some caution in using tools on peek fittings: if you break off a peek finger-tight fitting in a metal component, it's really tricky to get the plastic bit back out again. I've been reduced to getting difficult fittings out using Mole-grips, but always with a cold sweat.

[tom jupille]

it's really tricky to get the plastic bit back out again. I've been reduced to getting difficult fittings out using Mole-grips, but always with a cold sweat.

I've heated the tip of a smalls screwdriver on a hot plate and then jammed it into the stub of the broken fitting. Let it cool a bit and then simply unscrewed the stub.

[lmh]

Tom, thanks for that phenomenally helpful tip. I still reserve the right to be thoroughly unpleasant to the next person who breaks a peek fitting in a valco valve, but I'll try your approach before I let the engineers loose with a drill.

[tom jupille]

You're welcome. I'm living proof of the old adage that "experience gained is directly proportional to equipment ruined!"