Analytical & Environmental Chemistry Division

Looking for insight from anyone familiar with this instrument. I've recently started a new role and working with this instrument daily. I can't get an acceptable calibration curve for TKP. I've remade all reagents but my calibration keeps coming out awful.. points 3 and 4 are duplicated then 5 and 6 decline in value. It follows the same pattern each time I try. It's making me crazy and my boss (who used to be in this role) can't figure it out either. I haven't had this problem with any other methods, including other Phosphorus tests which use the same reagents (albeit in different concentrations). 

Help!

I see Seal's customer support for the AQ2 Discrete Analyser needs registration, which requires authentication.

A useful approach can be to system-split the process to establish where within the analytical process any issue(s) may lie.

Have you tried to deliberately run the same mid-range standard multiple times to see if the response is consistent?

Interestingly, when I Googled to see what I could find you'll never guess what the first search result was: -

https://www.google.com/search?q=discrete+analyzers+troubleshooting+curve

You mention the other phosphorus test doesn't exhibit this behaviour, what differentiates the TKP protocol that might be an issue?

Sorry Sandy, I don't have experience with this. I also asked my lab manager and he is not familiar with the equipment. What does it measure?

Ortho-phosphate following kjeldahl digestion. Our other ortho-phosphate protocols are all working fine, as was this protocol until a week ago. It's slightly maddening! 

Any developments?

I don't like to comment on assays with which I have no experience, but a couple of questions, Sandy. (1) Would it be possible to post an image of your defective calibration curve? (2) Have you contacted the company, its local agency or representative? 

@Thomas (Tom) Smith – My understanding, having exchanged a few ideas with her last week and checking in yesterday, is that @Sandy Lindsay has called in a technician to resolve what appears to be a liquid handling issue with the TKP method.

Only the TKP method on the discrete analyser was being problematic, which was partially diagnostic given the common reagents for other phosphorus determinations that were working. A two-fold reduction in the usual absorbance measurements at 880 nm indicated the reactions generating the molybdenum blue compound weren't occurring. Subsequently, it has been confirmed that the reagents work when solutions are manually mixed without the discrete analyzer.

It seems likely that there's a setting awry somewhere, which isn't immediately apparent from comparing the TKP method with other phosphorus methods. My latest suggestion was to make a copy of a working method and, so as to rule out a corrupted file, amend only those settings necessary to facilitate TKP.

I ran our top TKP standard (made in blank matrix) on another P protocol and got no result, below range of detection of the instrument. Even in a sample with a mismatched pH we would expect to see SOMETHING using the other protocol (though fully understanding the result would be bogus). So we could identify something awry with the matrix this way. Turns out the problem was the blank (acid) matrix..  The Nitric acid in the acid matrix seemed to be the culprit. Using a fresh bottle to make the blank matrix fixed the problem. (There is a half dozen open bottles in the lab and I have no idea which ones were used in previous attempts, so that's another problem to unravel)

I have no idea why I was able to get a result outside of the instrument, though I suspect it may just be because I scaled up the volume for ease of pipetting. Perhaps this helped with mixing?

That's great news, as at least you have identified the culprit.

Nitric acid is common to all solutions used to prepare the TKP calibration curve explains what you have seen.

Answering the question of why a particular batch or bottle of acid might be an issue is another matter entirely.

You're right that your being able to manually pipette the same solutions and get a result adds another layer.

At least you should now be able to perform TKP analyses using the discrete analyser.