We found the sensitivity(peak area) increses when the flow rate decreases. Can anyone explain why? Thanks.

Dear Anonymus;

The change in sensitivity you have seen is due to the fact that most LC detector are concentration dependent. In other words, they measure the concentration of sample in the mobile phase. The faster the fowrate, the less concentrated the samples are in the mobile phase.

Benjamin Esquivel

The main reason that peak area increases when you decreased the flow rate is because peak area is a function of time and signal. If the peak is wider in time (as is the case when the flow rate is decreased), the area will be greater. The area for detectors which are not concentration sensitive will only be constant when plotting response vs. volume.

Sensitivity isn't based on area, but on height. You may find that at lower flows, your peak height actually increases a bit due to the flow rate becoming closer to the optimum on the Van Deemter plot. Benjamin's response is actually the opposite of what actually happens w/ increasing flow - With faster flows, the "resistance to mass transfer" term of the Van Deemter plot increases more than the "longitudinal diffusion" term decreases, hence less concentrated (and therefore lower in pk ht) peaks.

I thought that sensitivity was the slope of the calibraion curve. So if you are calibrating on peak area wouldn't a lower flow produce greater areas and yield greater sensitivity.

Could you be looking at a detector time constant
effect? Short columns w high flow rates and early
eluting peaks and a slow time constant may cause
the problem.

No Name,
Calculate sensitivity as the signal to noise. Comparing the slope of a calibration plot works only when many conditions do not change. That is, compare apples to apples.
Jim

Check the chain "Calculating peak area via absorption coefficient" of May 16 (correct entry), there you get the mathematical basis for the relationship of flow rate and peak area.

On sensitivity: A long time ago when I was a student things were somewhat simpler. Sensitivity was defined via S/N, which everybody can understand. Here in Germany we now have four terms related to this, even though the math formulars behind the definitions are simple I donīt understand the reasoning. The confusion evident above is a result of these partially or totally unneeded proliferations.

The reason for observing the increase of sensitivity is explain very well by Benjamin and Chris. Some months ago there were some discussion about concentration vs. mass dependent detectors.

About "sensitivity" related considerations I find fair enough to have at least two kind of them: One is termed sensitivity which is the slope of the calibration curve and the other one is termed limit of detection (LOD) which is S/N related. While sensitivity should be about the same for the same instument in two different labs, the LOD could differ significally as backround noise can governed in relation of many factors (purity of solvents and additives used etc...). In this way you have the "potential" sensitivity that you can achieve and is instrument related and the other is the "experimental" one that will be affected by the lab practices.

Kostas

With the info that a nM of analyte gives a S/N of 4 (or whatever) you have the "experimental" as well as the actual "potential" sensitivity (or whatever) if it is stated that the optimum was achieved. The theoretical "potential" can be found in the manual or advertising. This theoretical "potential" is so trustworthy that it prompts many people to ask for a demonstration.
....Just donīt see it.