1
$\begingroup$

What would be the reasonable strategy for washing a reversed-phase (RP) octadecylsilyl (ODS) C18 HPLC columns from lipids and other non-polar residues? The columns are used for biochemical analysis of various matrices using acetonitrile, methanol and water (often with $~0.1\,\%$ formic acid), and it seems that lipids from cell membranes tend to cause clogging and increased pressure over time, despite the derivatized/extracted probes being filtered with 0.20 μm filters prior to injection. The process of washing with an excess (20× column volume) of 2-propanol recommended by manufacturer for removing lipid-soluble substances has only marginal effect, if any.

Usually, in this case it is recommended to try dichloromethane (DCM), dimethyl sulfoxide (DMSO) or tetrahydrofuran (THF) after replacing all plastic tubing and fittings with stainless steel ones. The problem is that the columns have polyether ether ketone (PEEK) lining that obviously cannot be switched, and all three solvents (DCM, DMSO, THF) are known for causing swelling in PEEK parts, possibly resulting in altered column geometry and traces of plasticizers.

Is there a reasonable solution for this issue involving another PEEK- and HPLC-friendly lipophilic solvent, or is there a totally different strategy for PEEK-lined columns?

Please note that I'm asking about the PEEK-compatible solvent in particular, not clogging issue in general:

  • The clogging occurs at the column, not anywhere else in the system. Replacing the problematic column with a brand new one or installing a resistance tube result in expected pressure values.

  • The retention times vary within acceptable limits. The columns are relatively new and were used for just several dozens of probes. There are, however, peaks of what appears to be post-eluting components despite a $\pu{40 min}$ washing cycle after each run.

  • Reversing flow direction and washing at low flow rates of about $\pu{0.1 ml min^-1}$ with either 2-propanol or acetonitrile didn't help at all. Although the pressure was a bit lower for the reversed flow (about $\pu{25 MPa}$ vs $\pu{30 MPa}$).

$\endgroup$

1 Answer 1

2
$\begingroup$

There is an dark side of HPLC with biological samples...with time, silica bonded column becomes a protein or a "bio-gunk" column, instead of a C18 column because of very strong adsorption of biomolecules on the surface. I gather your main problem is the pressure increase with these samples and the manufacturer's recommendation to wash with 2-propanol is not helping. You did not mention whether the retention time is also changing or not.

(0) The first test is to remove the column from the HPLC and connect the inlet and outlet tubings with a union. Start the flow and check the pressure of the system. If it is unusually high (>10 bar) at 1 mL/min chances are there is a cloggage in the system and the column is fine. Cloggage can occur in the injection system as well, just before the column.

(i) Run a QC test on this column with the manufacturer's suggested analytes and mobile phase. Are the retention times okay, within 0.2-0.4 min of the specified time? If not, this is indicating chemically changing silica surface.

(ii) Reverse the flow direction of the column. Connect the outlet side and make it the inlet. Do you see a significant change in the pressure? If yes, the inlet frit has some precipitated/strongly adsorbed stuff. Continue washing in this "opposite" direction for 3-4 hours and see if the pressure has dropped.

PEEK lined columns often have a PEEK or a titanium frit. With time, it happens that PEEK frit gets compressed due to pressure changes causing a permanent increase in back-pressure.

Edit: Try doing a step gradient of 95% H2O +5 % ACN to 95 % ACN + 5% H2O. Say 10 min of the first to suddenly switching the composition at a low flow rate say at 0.5 mL/min. Repeat this multiple times. ACN is pretty good for removing bio-stuff. Column beds hate viscosity shocks and they collapse fast. This could be tried at 0.4 mL/min overnight.

$\endgroup$
5
  • $\begingroup$ I edited the question addressing the points, of which we did pretty much everything already (except for QC — we need a new standard solution), that's why I'm asking specifically about the solvent, not the troubleshooting process in general. Thank you though for mentioning that there is a change of damaged plastic frit, that's something worth considering, although I don't know how to check its integrity. $\endgroup$
    – andselisk
    Nov 8, 2021 at 16:50
  • $\begingroup$ You can also add another clarification: Is your HPLC is compatible with normal phase solvents or not. The pump seal and the injector rotor may die after being exposed to DCM and /or THF. I will be more worried about the pump compatiblity because the column is disposable. Observing a change in the pressure by reversing the flow direction is indicating that the initial segment of the column is gooed including the frit. $\endgroup$
    – AChem
    Nov 8, 2021 at 17:17
  • $\begingroup$ Yes, there is no problem using chlorinated solvents or THF with the rest of the system, but we cannot change the column model because of SOP and method we have to follow. Again, the question is strictly about alternative PEEK-compatible lipophilic solvent or alternative washing procedure. $\endgroup$
    – andselisk
    Nov 8, 2021 at 17:19
  • 1
    $\begingroup$ Okay, as a last resort: Try doing a step gradient of 95% H2O +5 % ACN to 95 % ACN - 5% H2O. Say 10 min of the first to suddenly switching to 95% at a low flow rate say 0.5 mL/min. Repeat this multiple times. ACN is pretty good for removing bio stuff. Column beds hate viscosity shocks and they collapse fast. This could be tried at 0.4 mL/min overnight. $\endgroup$
    – AChem
    Nov 8, 2021 at 17:25
  • $\begingroup$ Oh, I think I heard of this method (something along the lines of "shockwave therapy"), but haven't tried it yet, only gradual switching back-and-forth between ACN and aqueous phase. Will give it shot, feel free to add this one to your answer) $\endgroup$
    – andselisk
    Nov 8, 2021 at 17:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.