By hplcgirl on Wednesday, September 11, 2002 - 01:02 pm:
Hi everyone:
Our lab has just moved and we are in the process of qualifying our instruments. I ran wavelength accuracy using an erbium perchlorate solution on a UV/Vis detector and a propylparaben linearity set for detector linearity per the waters protocol. A colleague disputes this qualification technique because it does not check the photometric accuracy of the detector. Who is right?
Thanks in advance for your comments.
By Anonymous on Wednesday, September 11, 2002 - 01:23 pm:
Two questions.
What does he mean by photometric accuracy?
Why does he care? Qualification is done to satisfy the needs of regulators. The procedure that you used has long met the needs of said regulators.
By Benjamin on Wednesday, September 11, 2002 - 01:50 pm:
Dear HPLCGIRL;
I am glad to know that someone uses the Er solution for wavelength accuracy testing. I developed that procedure many years ago.
Photometric accuracy of UV detectors is in a way implicit in the linearity tests. I know of no one who does specific testing of that sort. If your instrument had any problem reading absorbances, most likely there would be deviations in the linearity results.
If you want to be 150% sure the detector is working properly, ther are other tests you can perform, such as, noise levels in dynamic and static conditions, and RI sensitivity.
There is nothing writing in stone about instrument validation. Each laboratory is in a way free to establish how much validation is required.
Good luck;
Benjamin
By Ananda on Wednesday, September 11, 2002 - 03:05 pm:
When we moved our equipments to another location, we dealt with the same situation. We did not do the Er solution test but linearity test was done according to the Water's protocol using their test kit that contains methyl paraben. We were provided a 15mg uracil/L in MeOH instead of Er solution for the wavelength accuracy. So we basically did the same 2 tests. However, it is a common thing that once when you move detectors the alignments (horiz. & vert) can change and the reference energy of the lamp can be slightly off and that may affect the photometric accuracy. Still I also think this error should be detected in the wavelength accuracy test. That means the fact that he is concern is already solved by the 2 mthods that you performed. summary: he got a point BUT you are right!!!
By Anonymous on Thursday, September 12, 2002 - 11:56 am:
Even if you did perform a photometric accuracy test, what would you use for limits of the test. I do not recall any one manufacturer publishing a specification on photometric accuracy( of course, I am getting old and may just have forgotten). If you do not have a specification to from the manuf. to guide you, then you need to justify the limits you come up with. As you can see, it just opens a big can of worms doing this sort of thing. The tests you performed should adequately prive the detector is working. Maybe a noise and drift test also, but not photometric accuracy.
By hplcgirl on Thursday, September 12, 2002 - 01:14 pm:
P.S.
I was a Waters Engineer and have just come out of the field and back into the lab. I enjoy your forum tremendously and thank you for all of your responses.
By Tom Mizukami on Thursday, September 12, 2002 - 03:22 pm:
UV spectrophotometers are often used for direct quantitation. The extinction coefficient in a given media is known and the analyte is quantitated based on corrected absorbance. In this case photometric accuracy is directly related to the measurement error, and should be measured as part of the qualification procedure.
With HPLC, samples are normally quantitated against an internal or external standard so photometric accuracy is only indirectly related to measurement error. Like the others have said I think precision and linearity are more important for HPLC. Wavelength accuracy is critical if you want to quantitate based on relative response factors. Good Luck