Chromatography Forum

At my company we determinate RRF by the dividing least squares regression slope of the impurity by the least squares regression slope of the parent.

A question has come up regarding correcting for the purity of the material. Is it necessary to adjust for the purity to obtain an accurate RRF?


Thanks,
Robotjock

Those calibration plots are run with standards, right? If those standards are impure enough to matter for determining RRFs, then you have much greater problems to deal with!

Tom,
Thank you for your comment, but would appreciate additional explanation.

Typically, the RRF values are determined during method validation using the parent standard and a well characterized impurity lot. These values are then included in the official SOP. The impurity analysis described in the SOP uses the average parent RF results and the previously calculated RRF to determine the impurity wt% present in the test article.

It is realized if response factors are calculated directly i.e., peak response/sample concentration, then the material's purity can make a significant different.

However, there is some understanding(or mis-understanding) that use of the least squares process reduces/eliminates to correct for the purity.

thanks again,

Viewed from outside:

If you have a well-characterised standard that is 90% by mass Compound A and 10% by mass Contaminant B, and your most concentrated standard is 100ug/mL of dissolved stuff, then obviously this standard contains 90ug/mL of A; your calibration curve should have a point at 90ug/mL, not 100ug/mL.

Since your chromatography method must distinguish A and B, it doesn't matter how much B is in the standard when you are measuring A.

On the other hand, if you also need a calibration curve for B, since you know the standard contains 10ug/mL of B, you already have a calibration curve! You are therefore in a position to calculate the relative response factors. Hopefully your calibration curves will also be valid over sensible ranges, because you're using a standard that contains the two compounds in a realistic ratio.

Is this too far away from common-sense or good practice?

When using the linearity data to determine RRFs you will NOT need to take the purity of the compounds into account IF you already took it into account for constructing the linearity data .
In other words, for the linearity you're relating the analytical response (e.g. peak area) to the concentration of your analyte. Of course, the concentration of your analyte needs to be corrected for that analytes assay - no need to correct for a second time when calculating the RRFs.