removing ion-pair reagent contamination

jbarr

<p>We have recently completed the analysis of a set of samples using ion-pair chromatography. Although I was quite pleased with the results, the mobile phases we used, (nonafluoro pentanoic acid on one instrument, and pentafluoroheptanoic acid on another instrument - both are Aquity UPLCs), are proving extremely difficult to get rid of; both instruments show quite clear IPR parent ion contamination even after exhaustive flushing sequences. We've tried all the recommended flushing mixes (acid, base, strong solvent, weak solvent) and have also changed a number of different filters (solvent frits, in-line filters, mixer) and tubing (of course there is no column). In 15 years of working with LCs I have really never seen a contamination issue quite like this before; I was wondering if anybody had experienced a similar problem and could perhaps advise us on how to proceed. An online chromatography forum answer has just suggested that the degasser could be the source of the problem; is this a straightforward / cheap component to change over?

Kind Regards</p>

AJA

You should be able to troubleshoot the degasser without changing parts if you have a degasser channel that was not exposed to the IP agent. There are 4 solvent channels in the degasser, A1,A2,B1,B2. Assuming that you ran a A1,B1 method the A2, B2 channel should be uncontaminated. If you still see the IPR running on those channels, its not the degasser. If you do have to change it out, it is a fairly simple operation, maybe 20 minutes.

lizh

Hello:

Could send the method and configuration, which column manager and detectors and a list of parts already exchanged. We can supply a method and a list of parts to exchange further.

Also it might be useful to know which lines was the ion pair used in, for example did this include the sample manager washes?

We suspect that these compounds have a high affinity for PEEK, so anywhere that is present in the system is where you should start to look. Are you using a PEEK needle?

Many TX

Liz

jbarr

Hi, thanks ever so much for your quick reply. I've attached the methods; they were both methods that we received from Waters application chemists....nothing was really any different on our system. Unfortunately though, their flushing methods have not worked on our systems (Aquity UPLC-LCT Premier). Your comment on the PEEK tubing is interesting, since we have observed that on connecting the UPLC PEEK outlet tubing to a seperate reagent manager the contamination disappears, apparently ruling out this as the main source of the problem. However, we'll hapily change over as much PEEK as we can....I'm ok with the needle, but the other lines look more complicated to change over. I'm afraid A2/B2 were also used, in an attempt to improve peak shape on the early eluters (they were slightly better).

At the moment we have single column heaters on our systems.

Parts changed:

Plastic solvent lines and frits

in-line filters

mixer

Thanks again for your help

Jonathan

jbarr

ah sorry, that was an error; of course A2 and B2 were not used in the original IP method......unforrtunately though they would have been in the same bottles as the wash solvents that were used and would have been exposed to the IPR during priming......we routinely (biofluid metabolomics) have A2, weak = H2O and B2, strong = ACN, and I'm afraid I didn't specify to our technicians that this case would be any different. On one of the instruments (the method in the pdf file), the wash + A,B2 solvents were not exposed to IPR; I'll compare A1/B1 to A2/B2 on that instrument on Monday.

jbarr

Hi, thanks ever so much for your quick reply. I've attached the methods; they were both methods that we received from Waters application chemists....nothing was really any different on our system. Unfortunately though, their flushing methods have not worked on our systems (Aquity UPLC-LCT Premier). Your comment on the PEEK tubing is interesting, since we have observed that on connecting the UPLC PEEK outlet tubing to a separate reagent manager the contamination disappears, apparently ruling out this as the main source of the problem. However, we'll happily change over as much PEEK as we can....I'm ok with the needle, but the other lines look more complicated to change over? A2/B2 were also used, as were the needle wash lines. I'm afraid this is a rather regretful incident; we'd had several years of trouble free UPLC operation until now. We certainly won't be going back to this method and I would like to suggest that it should be removed from the Waters application notes list, at least for TOF detectors. I'd be very grateful if you could send me the recommended parts list.

At the moment we have single column heaters on our systems.

Parts changed:

Plastic solvent lines and frits

in-line filters

mixer

Thanks again for your help

Jonathan

lizh

Hello:

Based upon this post you appear to be seeing a constant background (this eliminates the Sample Manager and its needle and lines). It sounds like you have eliminated the MP with the reagent mgr test.

If this is the case Andy's suggestion about the degasser is a good one.

Can you describe your cleaning method - we may need a more aggressive version. For bioanalysis samples for example (with all the issues that brings along) we do recommend a good rigorous cleaning every 6 months to ensure the UPLC is clean and the MS probe is clean to restore sensitivity, so it may be that we should use this wash instead. I will attach some suggestions.

I have will contact the authors and get their recommendations as well.

Thank you

Liz

jbarr

Hello,

Thanks for your reply. One of the instruments also has the sample manager contaminated. Yes, the reagent manager test eliminates the MP (new bottles have been purchased anyway) and the MS system.

We used cleaning methods in accordance with your document; mixtures 1, 2 and 3 were used aas well as other MeOH, H2O, ACN combinations.....several litres at different flow rates and back pressures, over long periods of time; multiple full-loop injections of the cleaning mix were performed on the system with the SM contamination. I'm not sure if it'd be a good idea to use mixture 4. We are seeing a steady improvement, but this is not happening as quickly as I would like.

Please could you send me a full list of parts that you recommend we should be change (SM + BSM); it's not really a question of money at this stage, but getting rid of the IPR as soon as possible. We are due to move lab very soon.....I will ask the engineer responsible for the move to change the parts over for us, but we must get them ordered as soon as possible.

Thanks again,

Jonathan

lizh

Hello:

Working on getting the best possible method, but here are some key items to check.

1. Mixer
2. Sample Loop
3. Pre-Piercing Needle
4. Sample needle
5. Wash Station Block - Check the injection port and injection port o-ring for buildup of material, replace if necessary.

Our MS demo labs have experience with TEA contamination such as this, and washing for a very long time they have been succesful in removing the contamination, you will need to do frequent injections of the wash to ensure Sample Manager fluid path is clean too, (x10 full loop injections). They used 50% IPA:water containing 0.1% acetic acid for a long period of time.

Liz

lizh

I had forgotten about ASMS, the chemists that wrote the paper are away now travelling home.

Here is a more aggressive procedure that we have used for bioanalysis systems. It is similar to the 50% IPA wash, but throws in acid to help. This also breaks out the actual steps, so you can decide if you want to use the acid or not.

Liz

Steps for cleaning the UPLC

1.) Remove all solvent filters for the solvent lines

2.) Make sure the column is disconnected and connect the inlet and outlet lines with a union p/n 700002636

3.) Place all solvent lines; A1, A2, B1, B2, strong and weak needle wash lines, and seal lines, into a clean bottle containing 100% ISOPROPANOL

Note: If there is an incompatibility of the mobile phases /solvents on the system and 100% isopropanol,

flush with the appropriate solutions to ensure compatibility with the cleaning solutions.

4.) Prime the solvent lines ( A, A2, B1, B2) for 5 minutes for each line.

5.) Prime the seal wash

6.) Prime the wash and sample syringes for 5 cycles

7.) Purge the system at 0.2ml/min at 50% A1 50% B1 for 5 minutes

8.) Repeat the previous step using A2 and B2

9.) Perform 15 Full Loop injections (using 5X overfill) from a vial containing cleaning solvent. Use a run time of 0.5 minutes at 0.2ml/min

10.) Repeat steps 3-9 with 100% Methanol, except use a flow rate of 1.0ml/min for steps 7-9.

11.) Repeat steps 3-9 with 100% MilliQ water or equivalent, except use a flow rate of 1.0ml/min for steps 7-9.

12.) Take any MS detector out of line and connect with a union to waste.

Note: If a UV or PDA detector are in -line, they may also go through this cleaning regiment.

13.) Repeat steps 3-9 with 30% aqueous Phosphoric acid , except use a flow rate of 1.0ml/min for steps 7-9.

Note: Do not put the seal wash in this wash solution; leave it in the water until step 16 of this procedure.

****Note***** Never leave the ACQUITY UPLC in 30%phosphoric acid overnight**************

14.) Repeat steps 3-9 with 100% MilliQ water or equivalent, except use a flow rate of 1.0ml/min for steps 7-9.

15.) Place the detector back in line and proceed to step 16.

16.) Repeat steps 3-9 with 100% Methanol, except use a flow rate of 1.0ml/min for steps 7-9.

jbarr

Hello again:

Thanks for all the info.

We've more or less gone through all of this procedure already, apart from the acid step. Do you think it's significant? Remember that our detectors are high sensitivity TOF-MS systems.

Jonathan

lizh

That's why I said you decide! We do not want to swap one contamination for another one as you point out.

ANd its going to be a long wash. On Monday I will contact the scientists that developed the method. But definately look at the list of parts, do the wash and add in the new parts.

Best regards,

LIz

lizh

Jonathan:

Apologies, I got in contact with the chemist and they believe that along with the wash all the solvent lines need to be replaced.

This method comes from a poster and is used by a customer as a dedicated method and indeed suffers from contamination as a risk. When the application note is published this information will indeed be covered.

Liz