Chromatography Forum

Which is the most useful in determinig/optimizing lower detection limits?

In other words, should I modify my method for optimal peak height by adding a temperature ramp or would running isothermal be just as good? Are better detection limits obtained with a temperature ramp versus an isothermal method?

A similiar question applies to HPLC. Does running a gradient elution program tend to give low detection limits than an isocratic program?

Lower limits of detection depend on the ratio of the signal to the noise. Pushing the peaks through quicker makes them narrower and higher, which increases the signal. If the only source of noise is electronics and detector noise (which are largely constant with respect to column conditions), then a temperature programme, a high isothermal temperature, fast carrier flows, thin films and shorter columns all give the lower LOD. In fact the lowest LOD is provided by an uncoated column with no length - just inject the analyte to the detector in as narrow a band as possible. You can even stop the flow; put the analyte into a cuvette and let the detector detect it for as long as necessary. You will have spotted the problem with this - the analyte you are interested in is not separated from the rest of the sample, so your main source of noise is interference from other components.

With the kind of samples that chromatography is useful for you need to get a signal that is specific to your analyte, using either chromatographic separation or selective detection, and usually both. Sharp peaks gives lower LODs, but not if making them sharper compromises their separation from other components of the sample, which high temperatures and very fast gas flows often do. Also you will get more column bleed at higher temperatures, which increases the noise and lowers the signal:noise.

So the answer to your question is; it depends.

Peter

Without even talking about height or area, keep in mind that a ramp or gradient can accelerate the elution of later peaks, making it much easier to figure out where baseline begins and ends. And the abiltiy to detect baseline is important for both height and area.

Area sometime can be better. If you have a wide range in concentrations magnitude (butter fatty acids for example) for more concentrated peaks the base wide peak increase due to overload of column. In this way area peak mantain linearity while peak height not (height increase less than the right).

Robertino Barcarolo, Italy

And while sharper peaks are nearly always better, they don't do you any good if they are too narrow for the detector's response speed.

Peter

The peaks are well separated.

Would I get better detection limits with a temperature ramp or a long isothermal run?

In my experience, a ramp helps the later peaks - because they tend to be flatter. But that's on the stuff I've run. Best way to find out is to make some runs both ways and measure. Run a curve and see how the low end looks.

The peaks are well separated.

Would I get better detection limits with a temperature ramp or a long isothermal run?

Why not do a short isothermal run at a temperature near the temperature at which your substance elutes during a programmed run ?

All else being equal, the shorter the retention time the narrower the peak, the narrower the peak the taller it is, and the taller it is the easier it is to detect against baseline noise. So the critical call is not isothermal vs programmed, but long vs short. What is the fastest elution time that you can combine with baseline resolution from interferences ?.

For some compounds on some columns under you can elute sharper peaks by using temperature programming because the lower temperatures at the beginning of the run focus wide bands coming from the inlet.

So, ultimately, you need to make some injections and see what works best for the compound you are interested in.

Peter