Area vs. Height

[Stryder08]

What is recommended? We have a difference of opinion in my laboratory. Some think we should use height to quantitate, while others think that area under the curve is the better way to go. This is of course HPLC/UPLC with UV Detection.

[Uwe Neue]

Height depends on more experimental issues than area. Height can depend on the injection volume, the details of the sample composition (organic solvents, for example), column performance etc.. In addition, it depends on things that the area might depend on as well: flow rate, detector response, for example. Since there are less fundamental errors involved in peak area than peak height, I think the former is preferred. There can be additional issues with the integration of strongly tailing peaks, but in my opinion, this is a simpler issue to deal with.

[danko]

Hi Stryder,

In addition to Uwe’s argumentation I’d like to point out that in isocratic elution mode, the peak height is strongly dependent on the retention time. So, if everything else is equal (concentration, mobile phase, temperature etc.) the faster the compound elutes the higher is the peak. The longer retained peak/s will on the other hand be correspondingly wider, which will compensate the decrease of peak height.
This is especially important when quantitation is performed as a ratio to another peak. For instance related substances with different retention times are often calculated as a ratio to the main peak or the total area, under the assumption that the absorption coefficient for all peaks is equal.
The example above is a very strong argument in favor of using peak area over peak height.

Best Regards

[tom jupille]

Most data systems will let you do it either way. Why not run the method and compare the results using height and using area? If one clearly works better, use that. Otherwise, I'll join the chorus in singing "area is fundamentally preferred".

[unmgvar]

i agree with Uwe, Danko and Tom

in general using peak area is a lot more favorable then using peak height.
it is simply "more robust".

but there are some cases when using peak height might be more accurate for you for example if you are dealing with related compounds or impurity profiles where the resolution between your critical pair is very bad, and the height ration of those 2 peaks is significant.

there was a very nice work done by mbicking (you can find him on this forum), which was also published in the LCGC magazine:
"Integration Errors in Chromatographic Analysis, Part I: Peaks of Approximately Equal Size,“ M. K. L. Bicking, LCGC Magazine, 24,402 -414 (April, 2006).
–
"Integration Errors in Chromatographic Analysis, Part II: Large Peak Size Ratios,“ M. K. L. Bicking, LCGC Magazine, 24,604 -616 (June, 2006).

in the cases shown in those articles peak area can over-estimate some results by factors of 2 and more

[khw]

I also agree that peak area is preferred over peak height for quantification. However, I seem to remember an article (sorry, I could not find the reference), where for very small peaks quantification by height is more consistent vs by area.

[IC_Gal]

I can echo the above responses in that area is much more consistent than height. I have done trace analyses in IC for many years and the manufacturer even recommended that we calibrate by height rather than area in the low level work. A comparison of the two methods did prove this was so.

[HW Mueller]

Again, this has been discussed several times. I´ll dare to summarize and generalize: If the start and/or end of a peak can´t be determined but the apex is correctly "seen" by the electronics then peak hight is the better method. For other cases area determination is the method of choice.

In any case, one can test this fairly easily as Tom pointed out. One should be fair, though, calibration and performance consistency of the apparatus is even more critical with the hight method. If one is sloppy the hight method can misbehave badly.