Shimadzu UPLC pressure problem

[Lab_tech]

Hey all,

we use a Shimadzu Nexera X2 to determine banned azo-dyes in textiles. The column is 150/2 Phenylhexyl 1,8µm + guardcolumn . Our gradient uses 10mM phosphatebuffer at pH 6,9 and MeOH from 80:20 to 10:90 in 45min. Pressure should be ~380-400bar.
Now we have the following problem:
We rinse the system with H2O:MeOH 80:20 all time, switching to buffer:MeOH when running the method.(of course we eqilibrate before). When we switch from H20 to buffer, the pressure increases slowly over time (approx. 20-30bar per 30min) until the column/system max.pressure is reached.
Retentiontimes dont shift. It seems like the column head cloggs over time. Even if we change to a new column and run blanks, this happens too.
Shimadzu tested this method, after ~80 samples they had a slight pressure increase of ~50bar. They changed the guardcolumn an everything was fine again.

Any suggestions?

thanks

[tom jupille]

What you're seeing is consistent with the presence of fine particles and/or microbial growth in your buffer.

Buffers should be prepared fresh DAILY and filtered through 0.2 micron filters. Premixing the buffer with MeOH should be bacteriostatic, but it should be filtered after the addition of the MeOH in any case.

[Lab_tech]

thanks so far!

we prepare buffer freshly every day/ 2nd day. The buffer gets filtered through 0,2µm of course. Samples/standard/calibration solutions are filtered through 0,2µm too.

Last week we rinsed the buffer line with H2O, H2O: Isopropanol 50:50 and H2O:Isopropanol 80:20, then backwards to H2O.
Or what else we could do to eliminate microbial growth?

[tom jupille]

It sounds like you are doing things correctly, but you're getting junk from *somewhere*. If this were my problem, the next thing I would do would be to talk with Shimadzu to get their *exact* procedure and the check to see what I was doing differently.

[Klaus I.]

Is the filtration of the mobile phase really necessary? I have often observed such problems when the mobile phase was prepared with a filtration-step. Unfortunately I have no valid data about this. But IMNHO many of these strange phenomena would disappear when the filtration step is skipped. Assumed that appropriate solvents and buffer-salts in ULC or ULC/MS quality are used, I see no need for filtration of the mobile phase. I have got the feeling that filtration of such buffers (especially phosphate around pH 6) will catalyze the micro-growth.

Does flushing with 100 % isopropanol for several minutes (in the usual direction of the flow; Not back-flushing) of the column reduce the built up increased pressure?

[lmh]

just a thought, but have you ever used the line that you generally use for pumping methanol, to pump buffer? If so, depending on the make of solvent inlet filters that you use, it is just possible that you are getting buffer precipitation in the methanol line. I've seen this happen in systems being changed from ion chromatography to reverse phase, even after very extensive rinsing with water. I suspect that what happens is when we rinse inlet lines, the vast bulk of the solvent drawn through the inlet filter takes a route of least resistance, but somewhere away from this route, somewhere in the pores of the filter, traces of buffer remain. When we transfer the line to a phosphate-incompatible solvent, the solvent gradually creeps into all those pores, and buffer gradually creeps out, and when the two meet, buffer precipitates. Of course I'm far from certain this is what you're experiencing, it's just a thought.

[tom jupille]

Another thought: is there any chance that you are leaching something from the filters which is then deposited/precipitates on the column? Maybe a different type of filter will help (again, check with Shimadzu to see what they used).

[Lab_tech]

thanks for your replies!

Shimadzu sent me a restrictor capillary (sorry, dont know the proper english word) to check the system without column and some phosphate buffer solution. I'll test them next days.
I rinsed all lines with water then with isopropanol again, without column.

@tom, could be possible. shimadzu recommended those filters, so I think they use them too. We change them every 3 month, I'd say. Up to now Shimadzu hasn't any clue...

@Klaus, never heard of that, but might be a thought. the buffer substance we use has buffer grade for chromatography, by the way.
I never flushed the column with isopropanol. Only thing I did: flushing with water:MeOH 80:20, then increasing MeOH to 70%. Afterwards I ran water:ACN 20:80. Pressure decreased slightly after that procedure, but over time it increases again.

@lmh, we never changed the solvent lines. they were set up by a shimadzu technician as we bought the system.

Thanks for your ideas so far

[chrsdrake]

My first thought - are you getting slight precipitation of the phosphate at some point in your gradient? The phosphate concentration is so low this seems very unlikely. You could try running the initial conditions (80:20 methanol buffer) for many hours and see if the problem remains the same, you could try running repeated blank gradients to eliminate the possibility that it is a component in your samples (or autosampler) causing the problem.

Good luck

Chris

[Klaus I.]

thanks for your replies!
@Klaus, never heard of that, but might be a thought. the buffer substance we use has buffer grade for chromatography, by the way.

So the extra filtration of the mobile phase is most probably unnecessary and may add another source of particles?

I never flushed the column with isopropanol. Only thing I did: flushing with water:MeOH 80:20, then increasing MeOH to 70%. Afterwards I ran water:ACN 20:80. Pressure decreased slightly after that procedure, but over time it increases again.

I have observed this behavior in a few cases only and I have no valid conclusion for it. I think the test with isopropanol flushing would indicate that you have no problems with solid particles in your mobile phase (Therefore I mentioned, that it should be in the usual direction and not in inverse direction for back-flushing).

Maybe a more pronounced washing-step in your gradient will help? Go to 100 % Methanol at the end and hold these conditions for several minutes (5-10 minutes?). Please make before some experiments about the possible precipitation of your buffer salts and avoid rapidly changing gradient-mixtures.

Are you really sure that the increase of pressure occurs from the mobile phase (really without injection)?