[CLS]

Hi there,

I'm a technician in an analytical soil chemistry lab and I've been trying to think of a good reason for these confusing results; any help would be welcome. The instrument is an Elementar TOC/TN analyser, analyzing liquid samples extracted from soil.

Each run includes a standard curve made from a stock prepared in lab, a mid-run and end-of-run 'check' (just one of the points from the curve), a spike, and a QC - ordered from sigma. It analyses for carbon and nitrogen simultaneously with two sensors. Carbon is on an IR sensor, and nitrogen is assessed with an EC cell. Carbon isn't causing any issues, it's accurate and consistent run to run. The nitrogen, however, is giving me some grief.

Run to run, my standard curve has a very consistent slope. The spike recovery is pretty consistent and within +/- 10%. The in-run 'checks' are consistent and pretty accurate, so drift isn't a huge issue. But over the last couple months the QC has been slowly and consistently trending downward, returning a lower and lower value on each run.
I've gone through a couple of bottles of the QC solution and it seems to be decreasing the reported concentration independent of which bottle I use. I'll also note that I've used two batches of standard stock solution, and I make a new curve from the stock solution for every second run.

Ideas as to what might explain these strange results? I'm at a loss. I've thought about the possibility that the EC cell needs to be replaced, but then I would expect the slope of my standards to change or to become less linear. Same with the standards just being contaminated but, again, I'd expect a change in slope if that were the case.

Thanks in advance!

[marquis]

I used a different vendors TOC analyzer.  So this may not be true for yours. But the TOC worked by a UV lamp breaking down the TOCs in the compound  to CO2 that was then measured electrochemically. At higher levels, an oxidant was added that the UV light helped break down to CO2.  Anyway, it sounds like the UV bulb is dying.  Have you checked that out?    Good luck.

[CLS]

Thanks for your reply! Unfortunately, there is no UV bulb in my system.
I should have mentioned (I've edited the post now) that this is a combustion analysis; there is a combustion tube that sits at 850°C and injects the sample in to the top. The resulting gases are passed over a platinum catalyst, dried, and delivered to the sensors.

[CLS]

Thanks for your reply! Unfortunately, there is no UV bulb in my system.
I should have mentioned (I've edited the post now) that this is a combustion analysis;

Whoops. I'm new here and didn't realize that there is no option to edit posts. Hopefully this is an acceptable way of adding info.

[billnotgatez]

Thanks for your reply! Unfortunately, there is no UV bulb in my system.
I should have mentioned (I've edited the post now) that this is a combustion analysis;

Whoops. I'm new here and didn't realize that there is no option to edit posts. Hopefully this is an acceptable way of adding info.

You can only edit for a short time after you post to correct typos
BUT
forum rules say you should not edit after someone has posted a response
This is a forum policy.
Click on the link near the top center of the forum page.
Forum Rules: Read This Before Posting.
http://www.chemicalforums.com/index.php?topic=65859.0

[marquis]

Well, see if this makes sense.  How does the Platinum element look?  Is it worn or need cleaning?  Other than that, the temperature is the only thing that makes sense. Can the temp be checked other than by the instrument reading?    After that, the instrument company will have to help.  Good luck.