Any people ever used Excel to integrate LC peak area?

[yangz00g]

Is it possible? Thanks for any input.

[DR]

It's possible, but not practical. Just getting the raw data to fit in a single column can be a challenge depending on your RTxSpl Rate...

[yangz00g]

Thanks, DR.

I am trying a new software and it seems have some problems. I have some very nice peaks, but could not integrate them well (I am eager to see the results, so I want to kown if there is an instant alternative). The manufacture is working on it, and hopefuly they will get it fixed soon.

[tom jupille]

I've done it*, but as DR said, it's not a really practical solution.

Assuming you can export your chromatogram as a 2 X whatever array (time X voltage), then you can graph the plots to show the "chromatogram". Look at this plot to eyeball the start and end times of the peaks, then go back to your data array and do the following:

1. sum up the voltages between the start and end.
2. do a linear interpolation between the start and end points and sum those values.
3. subtract 2) from 1) to get the area.

This was a "quick and dirty", so I never saved the file.


* admittedly with "simulated" chromatograms.

[sksrinivas]

This is something new! I didn't know it was possible to do this on Excel.

However, I do remember a similar problem, in the days when I used a SP4290 integrator. The integrator had failed, so we plotted the chromatogram on a chart recorder (Varian 9176), at high chart speed and high full-scale deflection, so we got large, broad peaks. Then we simply sat and counted squares. Tedious and old-fashioned no doubt, but we were desperate.

If your software can print out your chromatogram with a grid in the background, perhaps you could do the same thing. Or, transfer it to Excel, print it out with a grid, and count.

Hope this helps!

SKS

[Rafael Chust]

I admit I never did it, but first generation PC data systems (20 years old or so. mainly from japanese manufacturers) had this export "feature", to be worked in Lotus 1-2-3 and QuattroPro...

Therefore, possible should be - pratical, I wouldn't bet!

[Rolandas Plausinaitis]

Hi,

We do it, but for one specific method (and we integrate not peaks but slices of the one big peak. (Actualy we import into Lotus 1-2-3).
There are two things you should consider even before starting work on that.
1. Baseline will be horizontal (you can not easily correct for the drift).
2. Calculated peak areas will not correspond to the numbers HPLC software is giving you.

[yangz00g]

Thanks for all the great suggestions...

I can imagine how "funny" it will be to count hundreds of small squares by hand