RPLC OF PYRETHROID: peak area reproducibility

[sofiane]

dear all,
I am using an RPLC method with C18 column eluted with acetonitrile-xater-isopropanol mixture, detection was set at 230 nm.

peak area is not reproducible, I consulted some articles and I have found that they have use an internal standard.
The official method is GC.
Is it possible to analyze the product with an external standard.
I note that linearity is not good
thank you

[tom jupille]

First of all, are you developing this method "de novo" or are you following an existing method?

Without knowing more details about *which* pyrethroids you are analyzing, your mobile phase seems to be somewhat complex. A quick search showed a separation of synthetic pyrethroids using methanol/water (https://www.agilent.com/cs/library/appl ... 3633EN.pdf)
and another with acetonitrile/water (https://academic.oup.com/chromsci/artic ... ermination).
The latter one in particular seems to give good reproducibility and linearity.

Whether internal standards will help or not depends on the dominant source of error in your analysis. In general, if the problem is primarily "upstream" (e.g., sample prep or injection volume), then internal standards will help. If the problem is primarily "downstream" (e.g., integration, baseline noise) then internal standards will make things worse. A quick-and-dirty way to tell (assuming there is more than one analyte) is to look at the ratio of the peak areas for two analytes. If that ratio is *more* reproducible than the areas themselves, that suggests the internal standards will help. If the ratio is *less* reproducible than the areas, then internal standards will not help.

[sofiane]

thank you,
the method is not 100% de novo, I follow existing method with some modification to improve specificity in my pesticide formulation and I find that no interference exist.
about mobile phase composition: it is composed by a mixture of acetonitrile and water with addition of 4 ml of isopropyl alcohol, according to Quality Control of Pesticide Products http://www-pub.iaea.org/MTCD/Publicatio ... 12_web.pdf
page 38 ' The addition of 2% 1-propanol to both mobile phase A and B had been reported to reduce the time required for re-equilibration at the end of the run and to improve column performance'
and page 118 : 'The use of 1-propanol in both eluents (A and B) helps the equilibration and stability of mobile phase, retention time will be very stable and reproducible'

because there is no 1-propanol in my laboratory, I use 2-propanol (decreaising time analyse and cost)

there is one analyte in the formulation and I have consulte the AOAC method for pesticide formulation (1990 ed), the most HPLC technique use an internal standard.

the formulation is an EC and I use an automatic injector. no traitement of the sample, dilution and injection.

I use the same conditions to calcule the purity of the matter and the amont of the pesticide in the formulation. last week I found 98.2% comparing to an standard (99.0%) and 83.3% for the formulation. later, I found 86% for the matter assay with freshly prepared solutions

note:
the injection volume is 15 µl and I have remarker that there is an 100 µl loop in the injector, I thik that the injection volume is controlled by the software

thank you again and I am sorry for my bad english

[mckrause]

I use 0.1% acetic acid in water:0.1% acetic acid in acetonitrile for all of the pyrethroids, using a 3 uM C-18 column (Restek Pinnacle). Gradient goes from 80:20 to 5:95. I have never had a re-equilibration issue and I get excellent separation.

I am working from memory, but I believe that 2-propanol absorbs UV at 230 nm. If this is the case you are going to have fits trying to get a method to work with a UV absorber in your eluant.

We don't normally use an internal standard for LC work, since LC's have sample loops. The internal standard is typically used to correct for injection variations.

[Rndirk]

Can you describe the problem in detail or show us a chromatogram? A peak area that is not reproducible: does that mean that you have a low signal, problems with background, interferences, lower signal in a subsequent injection? Is it for calibration standards or for samples only?

230nm is a rather short wavelength, as mckrause said you can have problems with solvent absorption + a higher chance for interferences in samples. Try to find or measure the UV spectrum of your component and measure at a longer wavelength. In the Agilent application note posted by Tom the detection is set at 270/300nm, it can make a big difference.

If your detector is a DAD, you can record the UV spectrum of the component.

[sofiane]

Thank you

I am working from memory, but I believe that 2-propanol absorbs UV at 230 nm. If this is the case you are going to have fits trying to get a method to work with a UV absorber in your eluant.

about isopropyl alcohol, the cutoff is 205 nm, and I work at 230 nm, the absorbance of the mobile phase will give an important background noise.

We don't normally use an internal standard for LC work, since LC's have sample loops. The internal standard is typically used to correct for injection variations.

this is the difference between LC and GC, but I have found in some AOAC HPLC methods for pesticide formulation the use of internal standerd (dialkyl phtalate for exemple)without pretreatment.

Can you describe the problem in detail or show us a chromatogram? A peak area that is not reproducible: does that mean that you have a low signal, problems with background, interferences, lower signal in a subsequent injection? Is it for calibration standards or for samples only?

I give you an exemple: peak area in not reproductible interday for freshly prepared solution comparing at the same concentrations using the same conditions.
the signal is better (about 0.3 au) and I verify the specificity of the method, no interferences exist.
The method is for standard assay and pesticide formulation. I compared the result of the same sample analysed in different days and there is 80% difference.

absorbance at 270 nm is lower than at 230 nm. I will show you a chromatogram and UV spectra later.
thank you

[tkubowicz]

Hello

First of all, if you have problems with peaks area RSD - check your sampler.
I'd recommend:

1.Inject 5 times the same standard (same solution, it could be from different vials)

It will help you to narrow down your problem. If RSD is poor (more than 1%) - diagnose sampler (rotor seal, loop, needle, needle seat, metering device). If RSD is less than 1% - investigate your sample prep.

Regards

Tomasz Kubowicz

[mckrause]

You are using an acidic eluant, right? Or at least buffered? If it is unbuffered you are going to have trouble.

[JI2002]

From Wikipedia:

"They are usually broken apart by sunlight and the atmosphere in one or two days, and do not significantly affect groundwater quality."

Could this be the reason for the poor reproducibility?