by
lmh » Wed Nov 23, 2022 11:09 am
I hate to disagree, but I disagree; it's nothing to do with air. You cannot pump air into a column with a UHPLC pump. The moment the pump head is full of air, all that happens is the pump piston squashes the (compressible) bubble and unsquashes it again, and nothing much moves anywhere. It's really hard to pump air into a column.
But if you've got a binary pump, and one channel runs dry, then you will probably be pumping something from the other channel. So, for example, if your aqueous solvent runs dry, you are now pumping a little bit of pure organic channel, and if you've got anything lingering in the system that can precipitate in 100% acetonitrile, it will.
Similarly, if you didn't have a lower pressure limit set on the pump (presumably you didn't, or it would have stopped pumping; setting a 5-bar lower limit can be a useful technique to guard against empty bottles), then if the bottle ran dry several injections before the end of the batch of samples, you've now made a a number of injections into a possibly unsuitable solvent, or a low/zero flow, meaning you might well have got a load of precipitated sample in the system...
So if you now have a very high back pressure, treat it as a typical blocked column and/or blocked autosampler (easy to check which) and wash with whatever procedures you know help.