A basic question in HPLC

[acquity]

Dear friends can anyone explain me why we do not prefer to use peak height instead of peak area for calibration in HPLC and other chromatographic techniques?

What are the drawbacks of using peak height?

Thanks right now.

[tom jupille]

There are times when peak height is preferred (e.g., very small peaks, or marginally resolved peaks).

Think of peak area as being the sum of a number of height (voltage) measurements.

[Uwe Neue]

On the other extreme: for well resolved peaks, the peak area is proportional to the mass injected, and is independent of the column plate count (i.e. column age) and injection conditions (extra-column bandspreading).

[acquity]

In our instrumental analysis class, our lecturer told that using peak height
is totally wrong. I am little bit confused now. If we thinkof the peaks as triangles, a triangle may have the same area with different heights.
The triaangle 1 will be the poorly resolved peak(lower height) and triangle 2 will be the well resolved peak.(higher height)
So far so good.

but if the resolution is same, then the height and the area will give the same result, won't it?

so why using peak height is totally wrong.

I will also discuss this point with my lecturer next lesson but I want to learn the right answer. He is telling a little bit wrong I think.

[DR]

Heights get more problematic in isocratic runs where height drops as a function of retention time (due to beak broadening, which is a function of diffusion, which is proportional to time). So, heights of peaks that contain the same amount of a given analyte change more with retention times than their corresponding areas do. Over the course of several runs using different mobile phase batches etc., heights will usually be more variable than areas on a per conc. unit basis of analyte, other things being equal.

In cases where resolution is marginal, especially when there is a large difference in ehights between adjacent peaks, heights can actually be more accurate.

[Uwe Neue]

The beauty about peak area is that it is simply proportional to the mass injected. As DR has pointed out, peak height will vary if retention varies, but it can also be affected by column age, injection solvents etc.

On the other hand, Tom has a good point too. If you have crowded chromatogram and need to quatitate an impurity peak on the tail of a parent peak, peak height might give you more consistent results than peak area.

I would go for peak area first, and use peak height only as the exception.

[tom jupille]

In our instrumental analysis class, our lecturer told that using peak height is totally wrong.

Never believe an absolute statement!

I am little bit confused now. If we thinkof the peaks as triangles, a triangle may have the same area with different heights.
The triaangle 1 will be the poorly resolved peak(lower height) and triangle 2 will be the well resolved peak.(higher height)
So far so good.

Actually, not so good.

First of all, resolution, by definition, applies to two peaks, not one, so that "poorly resolved" does not necessarily imply "wider". To a first approximation, peak width for a given compound under given separation conditions is constant and independent of the amount injected (lots of assumptions here, including not overloading, but a safe general statement). That means that peak height is, in fact, approximately proportional to the area.

but if the resolution is same, then the height and the area will give the same result, won't it?

Usually, yes.

so why using peak height is totally wrong.

As suggested by other posters, it's not totally wrong. It is, however, usually considered to be unwise. Re-read my earlier statement about peak width being constant, and note that I qualified it ("lots of assumptions . . ."). If you overload by injecting too large a sample, or if your column gradually deteriorates so that you lose efficiency (i.e., get wider peaks), or if your mobile phase pH drifts so that the peak begins to develop a "tail", or . . . and on and on, then peak height can change while area stays consistent. That's why area is generally preferred.

As DR and I pointed out, there are situations where peak height is preferred, but these are the exception rather than the rule.

I will also discuss this point with my lecturer next lesson but I want to learn the right answer. He is telling a little bit wrong I think.

Definitely discuss it with him (but be diplomatic!). I'm not sure about "wrong"; perhaps "oversimplified" would be more accurate.

[HW Mueller]

If one has complex matrices, like serum/plasma, the peak hight method might be the major one to use, because of the mentioned overlap. The shortcomings of this method are to be compensated by frequent recalibration (as discussed before, I mean here that one can run a standard after every 10th sample, or whatever). Let the calibration take care of the fact that the peaks get broader and lower as rt increases, etc. Also, the timid at heart can have their software determine both area and hight. A comparison will show very quickly which is better.