Chromatography Forum

Hello,
I am developing a method for determining the amidoamine (CAPA) present in cocoamidopropyl betaine (CAPB) via HPLC, using an ELSD detector, which does not yield linear results. However, I was able to obtain a calibration curve using the power curve function in ChemStation. When running unknown samples the software calculates the ppm based off the calibration curve. How do I take my sample weights and or dilutions of the unknowns in consideration when calculating the final concentration of the unknown? Or is the value that chemstation is calculating the true and accurate value based off area? When entering sample information prior to injection, it does allow me to enter sample weights and dilution, however it states that the sample weights are not used in reported calculation. After the result is given in (PPM) would I multiply by vol (mL) and divide by mass (g) to obtain a final final value or can this only be used if using linear calculations? Any assistance in regards to this matter will be greatly appreciated. Thanks in advance for your cooperation =)!

OK, if I understand correctly, you are currently using a one-point linear calibration. However the ELSD is only linear over a smaller range compared to UV-visible, and you know that. When we use ELSD, we try to adjust sample and standard concentrations so they're relatively close, like + or - 15% of each other.

Or, you can run a multi-point calibration, that's what that is for.

Thank you for your response, however I have performed a multiple point calibration using concentrations from ranges 1000-5000ppm for my standard. The correlation factor is ~0.998 using the %ESTD function provided by chemstation. The expected concentration in sample undiluted is ~3000ppm, after a 30% dilution i'm expecting at most ~900ppm. Just trying to understand where my sample size and dilutions play a role in calculating the final concentration in ppm

I think that the answer you are looking for is just to NOT enter the dilution info into the sample info screen. Don't worry about the non-linear curve fit with the ELSD. That is normal as they are usually linear over very small ranges. Your standards may cover a range where the response fits any number of "odd" curve fits. Which ever one results in the best correlation coefficient and is reproducible is fine.

*Your multi-point calibration table should be made up of a minimum of six standards which encompass your expected samples actual concentration value (ug/ml or whatever units you are appropriate). The standards should be made up individually to these value (not the same sample injected in different amounts on the column as this often induces error). If your calibration table is covering the correct concentration ranges, then your sample report will show the correct values and units for your actual samples and standards when you print it out. If it does not, then you not completed the calibration table info correctly.

Hi April,

It is possible to perform these calculations using the software developed by my company ChemTech Software Ltd. You can either do a specific or generic calibration, and our product FRAC will then calculate the concentration of your solutions both in mg/ml and mM units taking full account of the non-linerity of the ELSD response. It is possible to include dilution fctors in the calculations as well. The results are then exported in csv format for inclusion in reports, lab notebooks etc. This software is based on a publication in J Chrom A in May 2008 (http://dx.doi.org/10.1016/j.chroma.2007.11.017)

Please get in touch (Anthony.Squibb@chemtechsoftware.com) if you require further information.

Anthony Squibb

Head of Chemistry Products
ChemTech Software
http://www.chemtechsoftware.com