How to fix the peak area problm of calibration curve in HPLC

[GaussianWaves]

Hi everyone,

We're doing an HPLC analysis for 7 substances (beta-carbolines), and I'm using a multi-channel detection

For all substances, I'm using a 5-point calibration curve with doubled concentrations. The concentrations and channels for each analyte are as below:
Harmol (247 & 320 nm), harmane, norharmane (247 & 300 nm): .125, .25, .5, 1, and 2 ppm
Harmine (320 nm): 10, 20, 40, 80, 160 ppm
Harmaline (374 nm): 31.25, 62.5, 125, 250, 500 ppm
THH (223 & 290 nm): 2.5, 5, 10, 20, 40 ppm
Harmalol (374 nm): 10, 20, 40, 80 ppm

At 320 and 373 nm, peak areas doubled at the first concentrations but decreased at the third point, then the area slightly increased at the fourth calibration point and doubled at the fifth point, especially for harmine and harmaline. I'm not observing a similar problem at the standards at 223 nm, 247 nm, 290 nm, or 300 nm.

My injection volume was 20 μL, and I decreased it to 10 μL, but I'm still encountering this problem.
I'm preparing the calibration curves by mixing all standards with serial dilution. Therefore, it should not be a miscalculation problem.
I cannot find the reason behind the problem. Could anyone encounter a similar problem?
Thank you in advance!

[vmu]

If you use reference wavelength (360 nm?), switch this reference off.

[LuccasLN]

Hi GaussianWaves,

this honestly does not feel like a system problem to me, all molecules are somewhat similar, plain adsorption does not make sense because peaks would not double, system malfunction is not on the game because other WL work well and i respectfully disagree with the ref WL hypothesis since ref WL subtracts from the main chrom so again, we would not be seeing doubled peaks. My bet is on the sample prep, this has happened to me a few times in the past, specially if the analyte is hard to handle for some reason (ex. degrades, adsorbs, does not mix well,...). looking at it it seems like the analytes that are observed in 320 or 374 nm are in higher quantity as well in comparison to others, with exception to harmol... can you confirm this happens to harmol too?

How are you preparing the standards? do you prepare them individually or perform a serial dilution? make sure to make a serial dilution. also, it might me a little hard to explain here but while doing the dilution with an automatic pipette... pull for example, the mix for std5, gently pull and push the liquid that should be collected in the tip a few times, and when diluting it in the std4 reservoir don't just push the liquid... push it and then gently pull and push it a few times, this may help washing your pipette tip and avoid concentration fluctuations, make sure to mix it well before doing it again for the following standards. if it makes you feel more confortable you may discard your pipette every time youu prepare one standard.

you may also try more elegant solutions, for example, if you are really using an automatic pipette you have plastic tips, try to go for higher grade pipette tips or switch to glass pipettes (i know, not ideal, but we are just troubleshooting). same goes to reservoirs... once i discovered some of my vitamin E was sticking to plastic reservoirs (like those eppendorf centrifuge tubes) and i had better results by preparing my samples in glass test tubes.

also, while you are at it, do some testing with your pipettes, like take some water in different volumes and weight it. for example, you take 1 mL of water and drop it in a test tube, weight the water in it, it should weight 1 g. do this for the whole range your pipette is supposed to be working.

[Multidimensional]

This does sound like a classic training issue resulting in the incorrect use of the "Reference Wavelength" feature. Please turn Reference Wavelength OFF, by default on ALL methods. Next, re-validate the method and re-run all of the Calibration Levels to create new tables with the revised method.

"REFERENCE WAVELENGTHS (as used in HPLC UV/VIS):"; https://hplctips.blogspot.com/2011/03/r ... -hplc.html