Response Factor and Relative Response Factor

[jayroseville58]

I often encounter response factor or relative response factor on impurity calculation. When should I use those values and when should I ignore it on my calculation? I heard that as per E.P, if you have 0.8 to 1.2 RRF, the value will be just considered to be ignored or just equal to "1.0". Thanks

[tom jupille]

The response factor (RF) is the slope of the calibration line relating peak size (on the Y-axis) to the amount injected (on the X-axis). It is unique to a particular compound on a particular system on a particular day. The relative response factor (RRF), as one would expect from the name, is the ratio of the response factors for two compounds. In the case of impurities, it is usually RF of the impurity divided by RF of the parent compound.

- If you know that the RRF is 1, then both RFs are equal (obviously) and you can use the same calibration plot for both.
- 0.8 to 1.2 is approximately 1.
- QED

[jayroseville58]

The response factor (RF) is the slope of the calibration line relating peak size (on the Y-axis) to the amount injected (on the X-axis). It is unique to a particular compound on a particular system on a particular day. The relative response factor (RRF), as one would expect from the name, is the ratio of the response factors for two compounds. In the case of impurities, it is usually RF of the impurity divided by RF of the parent compound.

- If you know that the RRF is 1, then both RFs are equal (obviously) and you can use the same calibration plot for both.
- 0.8 to 1.2 is approximately 1.
- QED

thanks again Tom for the reply, so it means, it is right that I should always assume that 0.8 to 1.2 RRF is equal to 1.0?

[tom jupille]

it is right that I should always assume that 0.8 to 1.2 RRF is equal to 1.0?

No, only if that's what the method states. What they are saying (although in an awkward way) is that they will tolerate an uncertainty of +/- 20%.

[gtma]

Note: When the RRF is very small (e.g. < 0.5) or very large >1.5 there is a much higher probability for observing significant variability in the impurities results from run-to-run. Therefore, one should avoid using RRF that's too small or too large.

On a different note, if your LOQ for several impurities are different due to differences in the RRF e.g. 0.02%, 0.05%, 0.08%, 0.10% . Should you consider the method LOQ to be the 0.10% (worst case) or should you apply the corresponding LOQ for each impurities? What about unknowns.......use the LOQ for the API peak?