plate number calculation

[pph31]

If I use diluted sample which gave lower peak height but narrow half peak width, the plate number will be higher. Are I manipulating data?

thanks

[scottythree]

Are you using a validated method and are you referring to diluting a standard as opposed to a sample? I am used to performing plate counts for system suitability checks using standards only (with a known purity and concentration) as samples can be very variable.

[pph31]

same std, but inject less which leads to smaller, narrower peak.
method used for calculating Plates should be fixed. no change including std conc, injectin volume.

[tom jupille]

If you are not overloading, the peak width at half height should be approximately independent of the amount injected. And, if the width *does* increase with increasing amount, then you *are* overloading.

[pph31]

If are not overloading, the peak width at half height should be approximately independent of the amount injected. And, if the width *does* increase with increasing amount, then you *are* overloading.

are you saying the peak width of 10% and 50% std are the same if the method is validated and not overload?

[tom jupille]

Yes, the peak width should be independent of concentration so long as you are not overloading. The peak height will be proportional to the amount injected. That is why peak height can be used for quantitation; width is constant, so area is proportional to height.

[pph31]

Yes, the peak width should be independent of concentration so long as you are not overloading. The peak height will be proportional to the amount injected. That is why peak height can be used for quantitation; width is constant, so area is proportional to height.

I will look into my data again, will confirm it tommorrow.

Thanks, it helps.

[pph31]

Yes, the peak width should be independent of concentration so long as you are not overloading. The peak height will be proportional to the amount injected. That is why peak height can be used for quantitation; width is constant, so area is proportional to height.

I will look into my data again, will confirm it tommorrow.

Thanks, it helps.

The peak half width was different from 1 to 120% std. Their response factors are same using peak area/conc.

[tom jupille]

That means that your system is overloaded (as far as the chromatography goes). While that is not the ideal situation, it is not necessarily a problem. I does mean that plate number and resolution measurements should be made under "worst-case" conditions (i.e., with the largest standard in the linear range).

[HW Mueller]

One can also suspect that the halfwidth measurements of the lower peaks are off. If no mistakes, like overloading, are perpetrated than all peaks should overlap almost perfectly if you normalize their hights to be the same in an overlay mode of the PC software.

[Peter Apps]

This can also happen if the large peaks are out of the linear range of the detector - the top of the peak is blunted, the peak apex is lower, and the half height is lower down the sides of the peak where the width is greater. This shows up clearly if you follow Hans' normalize and overlay suggestion.

Peter