BEH column lifetimes

<p><span>has anyone else had problems with BEH column lifetimes? We're only getting a few hundred injections on ours before they void, Waters have said its down to our method conditions (pH10 bicarb/MeCN gradient @ 45°C) dissolving the stationary phase... </span></p><p></p><p><span>We've changed our conditions to 30°C now and are seeing if it makes a difference, I just wondered if this problem was exclusively ours... <img alt="" border="0" src="http://forums.waters.com/clearspace_community/images/emoticons/confused.gif"/> </span></p>

Answers

  • Nope, the problem isn't just yours. I've got a method for estrogen compounds using an ACN/H2O gradient @ 30° that has gone through three columns in less than 1000 injections each (two were less than 800) before pressure increases and sensitivity decreases. Conversely, I have a method for sulfonamides using a MeOH/H2O + 0.2% Formic gradient @ 40° and the column I'm using for that method is rapidly nearing the 3000 injection mark. Whether this difference is due to the mobile phase being used or the compounds being analyzed, I don't know.

  • peakdalejo, have you confirmed the presence of a void by opening up the inlet of column and looking at the bed? If so, may I ask what your maximum pressure is in your method? That data should be on your eCord. What is the concentration of your ammonium bicarbonate buffer? Do you record the actual pH of mobile phase A? Have you ever run a set of standards/blanks to see what the column lifetime is? Apologies for all the questions.

    If you've not opened up a column that you suspect is voided, I suggest you do so. It's quite easy to do. If the packed bed is NOT voided and the packing material is to the top of the inlet of the column tube, discard the old inlet frit, place a new inlet frit (you can use one of the in-line filters, assuming we're talking about a 2.1 mm ID UPLC column) into the inlet nut, re-attach the nut/frit, tighten and re-test.

    Basically it is impossible to mechanically void one of our UPLC columns because of how they are manufactured. Dissolving the packed bed is the only way to do it. High temps + high pressure + high pH CAN do it in some instances, but it is quite rare and unusual.

    Thanks,

    --Doug

    Edited by: Doug McCabe on Jan 8, 2009

  • Hi Doug, yes we have opened one of the columns and its definately voided! Pat Boyce from Waters has taken a look at it too and we took this photo (attached) as he was quite shocked to see it!

    The max pressure of the method we use is about 11500 psi and we adjust the pH of the bicarb to the range 9.95-10.05 with a pH meter (which we calibrate before use), its concentration is 10mM. I've not tried just running blanks and standards to get a lifetime but its generally no more than 700 injections at present. Its particularly annoying as the first couple of columns we had for this system were fantastic, lasting for >3000 injections, one even made it to >9000, but recently they've been very disappointing. We also use XBridge on our HPLC systems with the same methods, these have always run at 30°C, and they really are indestructable!

    Pat has recommended we reduce the temperture to 30°C on the UPLC so its directly comparible to the XBridge methods. I've done this and adjusted the flow and gradients to give similar chromatography. Hopefully this'll make the difference...

    Hoth, it sounds like you're having particulate contamination problems, the pressure on my columns remains fine but my peaks start to tail and then split due to the voiding.

  • I was wandering if going to 30 deg C solved your problem. I use similar conditions (10 mM ammonium bicarbonate, pH 9.5 / Acetronitrile. Column temp = 60 deg C) and am experiencing poor peak shapes after 500-600 injections.

  • Hello,

    Have you experienced shorter lifetimes on multiple BEH columns under these conditions? Have you always experienced the same lifetimes or was there a sudden decrease?

    Just trying to understand your particular situation a bit better.

    In general, for any LC column, if you lower the temperature when working at high pH and high pressure, you can extend the column's lifetime, assuming the cause of column failure is phase dissolution/hydrolysis.

    --Doug

  • Yes, I have gone through 5 columns since I have started using this mobile phase system. None of the columns have lasted for more than 700 injections, and severely assymetrical peaks were observed during the last 200-300 injections. Indeed, I am using a high temperature (60 deg C), and pressure is typically just below 8000 psi for a new column, which peaks to just over 10000 psi during the gradient. The primary reason I am using this solvent system is to achieve sufficient sensitivity, which is a factor 10 better than with acidic solvent. With acidic solvent, however, the retention times are more stable, peaks are more symmetrical and my collegues using acidic solvents have not had any problems with column degradation so far. I have not opened up a column, but extensive washing with apolar solvents does not help. A further disadvantage of this solvent system is that the decomposition/volatility of ammonium carbonate seems to have a pronounced effect on my retention times, with peaks shifting by as much as a minute over the course of a long run of samples (over the course of a weekend)

    I am going to try using milder conditions. Which will have the most effect: lower pH, lower temperature or lower pressure?

    Reducing the temperature will increase the pressure, so do you think the overal effect will be beneficial? (I can't reduce the flow by too much, since I onyl achieve baseline seperation of my components at a flow of 0.6 ml/min)

  • Hello,

    The one parameter that will likely have the most beneficial effect while having the smallest effect on your chromatography is lowering the temperature. I am not suggesting the most obvious thing that will increase column lifetime (lowering the pH) since that means that you have to redevelop your method and/or you will have a method that is inferior to what you have now.

    Yes, at lower temps you will increase your operating backpressure if you keep the flow rate the same. You've already did a pretty good job diagnosing the problem since you and your colleagues do not have the column lifetime issues when operating at low pH.

    The RT change that you are seeing over the course of long runs is most likely due to the buffer absorbing carbon dioxide from the air, resulting in a lower pH. Are the basic compounds RTs decreasing while the neutral compounds RTs remain unchanged? To confirm CO2 absorption, you could check the pH of your buffer at the end of your long run.

    Good luck and please let us know how it goes.

    --Doug

  • Hello,

    Have you tried the Waters Vanguard pre-columns at all?

    I find them quite useful in extending the life of my BEH C8 2.1 x 50mm columns. (I normally run at pH 6.8).

    If performance deteriorates you can just fit a new one. Sometimes it works, sometimes it doesnt.

    Also, I regularly try a regeneration method if peak shapes start to deteriorate - I am injecting 5 X 10ul DMSO injections (- containing 20% Formic Acid), and running a gradient at 70 -100% ACN (B line) over 1.5mins at 1ml/min (A is 0.1% Formic in Water).


    This seems to shift (hydrophobic?) material that may have built up at the column head - or al least thats my assumption.

    Have regenerated a column with poor peak shapes about 10 times now - seems to works well!.

    Cheers,

    A.

  • I know it has been quite awhile since you had posted this, but I was wondering if you are still having this issue? I am also having split peaks with a estrogen compound method using 50:50 ACN:water @ 40C. Split peaks were observed at ~800 inj. with a 25% EtOH / 5% PG sample diluent. It was suggested that the PG was causing contamination. The sample diluent was changed to saline and on the first column I ran this on the split peaks were observed at ~250 inj.

    I was just wondering if you were able to adjust your method to increase column life or if you found what was causing your issues.

    Thanks,

    Rob

  • Hello RHuber,

    I dony know if you saw my post the other day, but the DMSO/Formic method I use regularily eliminated split peaks for me. Of course it may not work fo you, but it might be worth a try.

    Cheers,

    A.

  • I do use the vanguard precolumns, but fitting a new precolumn did not resolve the issue, unfortunately.

    I am going to try regenerating it with DMSO injections as per your method and report back.

    The manual describing column care and use does mention it. The same manual also suggests flushing with 7M ureum or a sequence of acetonitril/isopropanol/THF/hexane, neither of which yielded any improvement.

  • I don't think the problem is related to pH as we have been having the same problem with samples run precipitated with 0.1% formic acid which is also in the solvents, always run at 40C.  We had used the same column for almost 2 years getting about 2000 samples/column.  When the columns changed to the yellow cap on the e-cord magnet, the lifetime became more variable and in some case, downright horrible.  Over the last year we have been through at least 15-20 columns with some losing peak shape and having much shorter retention time within 300 injections and others lasting 1500-2000 injections with decent peak shape.  Pressure isn't an issue in the columns that don't last as long as it's only 7000 or so vs. other columns which still give good peaks at 11000.  I was told that the new columns were supposed to be improved but performance and consistently apparently were not part of the improvements.  I haven't opened a column to see if there is increased void space nor have I tried DMSO injection, but now I'm curious to see if that changes anything.

  • In the end we changed to the 2.5um XBridge column in the same dimensions after we couldn't get lifetime form BEH.  The small change in particle size made a big difference in lifetimes for us.  It went foorm around 1000 injections to over 3000.  There was a little loss in resolution but not enough for us to worry about and we have not adjusted the methods at all, just literally swapped the columns over and carried on running.

    In then end we came to the conclusion that the combination of high pH, high pressure and fast gradient (we're running 2-98% MeCN in 0.8min) was just dissolving the packing material.  We think the 2.5um XBridge lasts longer as it takes longer to dissolve...

    I'm not sure what implications changing the column would have in your methods but it may be worth a go?

    Interestingly we've been steadily testing and destroying other manufactureres 'equivalent' phases and none have come close to 2.5um XBridge for the lifetimes using our method until now.  We've been trying the new YMC Triart column and its the only one we've tried that has matched XBridge as far as lifetimes go, performance is directly comparable and we've yet to kill it, I think its up to around 3000 injections at the moment.  Once we've taken a closer look at it as far as scale up to prep and all round performance goes we may have finally found an alternative column for our lab which runs 90% of the compounds we have in at high pH.

    Another interesting point we have noticed with XBridge 5um on our other HPLC systems is that a column that appears to have died - poor peakshapes, if stored in 50/50 water/MeCN for a month or so will come back to life and carry on working for another 6 months or so!  We have saved hundreds if not thousands of pounds by saving our 'old' columns and storing them for a few months before retesting them and putting them back in to circulation if they are ok.  The theory on this is where we run almost exclusively gradient methods we eventually compress the column bed and by letting it relax back for a period of time they recover of their own accord...

  • Hello Cheryl,

    If you wish, you can contact me directly in hopes of improving your UPLC column lifetime. My email is Doug_McCabe@Waters.com.

    We strive to apply product improvements that not only produce better performance, but more consistent performance. As a result of your posting, I would like to work with you in order to get your column lifetimes back on track.

    I look forward to hearing from you.

    Best regards,

    --Doug