PDA Calibration

[HPLCSlave]

Hi!

I would like to ask whether somebody can explain which phenomena drives to the need of calibrating a PhotoDiodeArray(PDA) detector.

Thank you very much, in advance!

[tom jupille]

Calibrating wavelength or absorbance?

[HPLCSlave]

Oh... This is quite a difficult question... Would you mind if I ask what is the difference between the 2 calibration types and their use?

Thank you very much!

Best regards!

[tom jupille]

That sounds like an examination question (which I won't answer for you). I can give you an analogy. Imagine you are assigned to produce a map of rainfall in the Rhine valley. The wavelength is analogous to position on the map. The absorbance is analogous to the rainfall level.

[HPLCSlave]

Ok... I have decided to use a satellite mapping system for my map; I must calibrate the satellite so as to correlate its distance unit to the real one on the
earth. I take 2 big objects and I measure the distance between them and I explain the satellite that between the 2 houses, let's say, the distance is exactly 22 Meters.
Now, for the rainfall level, I use some special bottles . They have an emitter, 2 valves and a weight detector(scale), assuming that water density is always the same:
1. valve for letting water in but not allowing it to evaporate
2. valve for releasing all water out when the bottle is full
3. emitter providing the following information:
a. an ID of the bottle corresponding to its position on the map
b. time elapsed since last remote controlled water evacuation
c. the weight of the acumulated water

I go to the Standards Institute to ensure that the bottles have the same volume and that they are made of a material that have a small dilatation coefficient, for the measurements must be correct in winter as in summer.

And, finally, I sit confortably on a sofa and I connect all bottles to my computer and observe how much it rains as time goes by...


But how does the PDA-UV detector calibrate??? I only press a button and it is claimed that it calibrates itself! Wavelenght? Absorption? ( I suppose absorption) but how?
Where does it have a standard?
I was just curious...
Is it so for the absorption that it closes the detector and this means absorption is zero?

[DR]

Calibration is probably done w/internal sealed Holmium oxide cells - consult the DAD's manual to be certain. If the manual isn't near the instrument, it's probably available online from the vendor's web site.

[tom jupille]

That was waaay more detail than I intended on the analogy!

What any UV detector does is to pass a beam of light of a specified wavelength through a flow cell which contains the effluent from the column, and measures how much of the light gets through. So, there are two things to calibrate:
- the wavelength (are you sure that you are actually using the wavelength you think you are using?) which is analogous to position on the map, and
- the absorbance (is the detector accurately measuring how much gets through?), which is analogous to the rainfall.

Photodiode array detectors and "conventional" variable-wavelength detectors arrange the optics a bit differently (there is a summary in the FAQ section: http://www.lcresources.com/wiki/index.p ... d_a_PDA.3F ) but basically do the same thing.

Absorbance accuracy is generally not a major concern in HPLC, because quantitation is always based on comparison with known standards run on the same instrument. If the absorbance is off by, say, 10%, the effect will be proportional on both the standards and the samples, and so will cancel. If you really want to calibrate, you must fill the cell with a solution of known absorbance and compare the detector output reading with the known value.

Wavelength accuracy is somewhat more important, but even there, not as important for HPLC as it would be for spectrophotometry. Proposed standards for method adjustment from the USP and the US FDA, for example, allow +/- 3 nm error. UV spectra in solution (unlike, say IR or vapor-phase UV spectra) are information-poor (essentially broad lumps), so that sample identification by matching spectra is questionable at best. To calibrate, you must fill the cell with a solution of a compound with a known spectrum and compare the wavelengths at which the absorbance maxima occur.

As DR stated, instruments with built-in calibration usually interpose a filter of known characteristics instead of filling the flow cell with a solution. This is much better for wavelength accuracy because a solid filter will have a lot more "structure" in the UV spectrum.

[HPLCSlave]

Thanks for your kind help!