carbon/ biological contamination of HPLC?

[dkreller]

Hello,

Thanks in significant part to the input from people on this forum, we have successfully implemented a HPLC method (size exclusion) that has an online continuous-mode dissolved organic carbon (DOC) detector in our lab. This detector is great because it sees everything that has carbon in it, whether it absorbs light or not. The problem with this detector is that is sees everything that has carbon in it.....

For a couple of months we had very low DOC baselines, meaning very low DOC bleed in our sytem. During this period we got some good data. We had played around with different columns as part of our set-up phase. However, currently we are having a problem in which the baseline of the DOC detector is unusually high and variable. One possible cause of these problems is the fact that for about a week we were trying to set up a reverse-phase method on the HPLC and had a lot of MeOH pumping through it. However, I have pumped pure DW through the system for days and days to try to flush any residual MeOH out, and that hasn't solved the problem.

The DOC of the effluent fluctuates, which indicates that a reservoir isn't contaminated. I am wondering if we have had some microorganisms colonize somewhere in our system. Has anyone dealt with/ worried with that sort of thing? If so, what measures do you take to control 'bugs'? Flush with an azide solution? On the advice of Waters corp. I tried to flush (without column) with 6N HNO3 for 60 minutes. That didn't solve the problem either.

Any input? Thanks in advance,

Dave Kreller

[MK]

intresting..

Atleast TOC analyzers has a special "clean mode" where low TOC water is pumped through the system while UV lamp is on to oxidase all organics to CO2 and H2O. I would imagine elevated temperatures and reverse flushing could remove more dirt.

You could also look at how glassware etc. is cleaned for TOC sample handling.

[HW Mueller]

Have you checked your mobile phase taking, successively, the column, Rheodine, +? off line? No carbon left?
I just had some problem with a drifting baseline in UV detection and found that part was due to material diffusing out of the Rheodyne.

[dkreller]

Hello,

Our DOC analyzer is not the issue. When the HPLC effluent bypasses it, its DOC reading falls to the ppb range.

A wierd thing is that the contamination seems to be the worst in a specific line. When I pump pure DI through the system, the DOC baseline falls to a relatively reasonable level, but then when I try to add another mobile phase component, a salt solution from a different reservoir that gets diluted in the DI, the DOC baseline jumps up substantially. The solution in the reservoir is not the issue, with a separate batch DOC analysis (Shimadzu) I have seen that the solution has ~0 ppm carbon.

Perhaps I can try to clean that specific line with HNO3.

Again I will put this question out there. Any suggestions on how to keep microorganisms out of an HPLC system?

Dave

[HW Mueller]

You apparently misunderstood me, I was not talking about the DOC, but rather about isolating all other parts, like the column, from the plumbing. Actually that´s what you appear to be doing.
I don´t want to discourage you, just extol you to be careful: Years ago we gave up on an analysis of oxalic acid via elemental MS (also via CO2). The problem was, of course, that we could not get rid of extraeneous carbon on the way to CO2.
HPLC/GCMS of the whole molecule was getting somewhere when interest (by physicians) waned, so we stopped that also.

[Chris Pohl]

dkreller,

Do you have a low pressure gradient HPLC pump? If so, have you rinsed out all solvent lines? You might be surprised to see that the metanol has gotten into parts of the system you hadn't expected. When faced with a similar "incompatible eluent residue" problem I do the following: rinse out all eluent containers 2-3 times with DI H2O, discard the endline eluent filters if possible and repalce with fresh ones if possible, flush all eluent lines simultaneously while throwing the injection valve repeatedly. Repeat the rinse out all eluent containers 2-3 times with DI H2O and again flush all eluent lines simultaneously while throwing the injection valve repeatedly. This should bring down the background problem unless it's column related. What column are you using?

[dkreller]

Chris,

Sorry I have not been on this forum for a couple of months. I am hoping you will still check this. We definitely do hardcore cleaning of the glassware including 10% HCl bath and oven heating to 450C. Thanks for your suggestion of changing the filters. However, I don't think any particular line or component of a particular line is the source of the contamination. I observed that when DI water was passing through any of the lines, the DOC reading would fall to a reasonably low level, but when I SWITCHED the flow from one reservoir to another there would be a spike, almost looking like a sample injection, in the DOC a few minutes later. Then the DOC would fall back down and whenever I made such a change in the flow the same spike would appear again. This made me think that the gradient proportioning unit that was the source of the most contamination, and when it gets used a slug of gunk falls off it. I did a serious 'passivation' of the system with first 6N HNO3 and then MeOH aimed at flushing each of the lines and cleaning the best I could the gradient proportioning unit. Our DOC analyzer was down for a week and is now working again, so I am picking up where I left off - seeing if this recent flush/ cleaning made a difference.

We are using a TSK column (G3000PWxl). We switched to TSK for the DOC application because the Waters column we were using was showing some constant bleed of several ppm DOC and making the baseline high and variable.

thanks for any further consideration,
Dave

[Chris Pohl]

dkreller,

Well, there is still a possibility of hideout in several different places in your system. If you're sure you have all of your eluent lines cleaned out, you might still want to try operating with high purity water in all eluent containers at the same time with the percentage of each being split evenly between the eluent containers (e.g. 25% of each in the case when you have a 4 eluent proportioning valve). If that still doesn't resolve the matter, try bypassing the proportioning valve altogether. Does this help? If so there may be some sort of mechanical failure in the valve and it might need to be replaced. Another possible problem could be a contaminated online degassing unit. If you're pump has one, and the previous hasn't been any help, try bypassing the degassing unit to see if that helps. Another possible hideout area is in the pump heads. If you stay with a single eluent source and get low backgrounds when you start with a fresh batch of clean water in a given eluent container, what happens if you stop the flow a few hours and then resume flow? A big jump in contamination suggests contamination from poorly swept regions of your instrument such as the poorly swept areas around the check valve cartridges. You might want to disassemble the pump heads, sonicate everything in high purity water and reassemble them. Another somewhat unlikely contamination source is via cross-contamination through shared eluent bottle pressurization lines. I've only seen a bona fide example of this once (two instruments sharing the same cylinder of helium where one of the pressurized eluent containers from one of the instruments had one molar ammonium hydroxide which caused high ammonia blanks in a second instrument which also had an acidic eluent bottle connected to the same cylinder). Perhaps this could happen if you are working with pressurized eluent containers with a shared gas pressurization cylinder with one of the eluent bottles on a separate instrument containing a volatile organic solvent.

[dkreller]

Hi Chris,

Wow, you must have some extensive experience in tearing apart HPLCs. I think you may be really hitting 'the nail on the head'. Thanks for the very good critical thinking on this. Thank you.

Dave