Determining response factors

[MarkFG]

What is the way to determine response factors by HPLC for an API and its impurities?

[sassman]

Response factor is just instrument response (usually peak area) divided by analyte concentration. Do you have standards for the impurities?

[bartjoosen]

Response factor is just instrument response (usually peak area) divided by analyte concentration. Do you have standards for the impurities?

one more question:
Analyte concentration as mol/l or g/l?

[MarkFG]

I have the API and its impurities. I have analyzed them all at one concentration (mg/mL) to correlate retention times in both the API and a product formulation. However, I only know the purity of two of the six impurities.

[DR]

Getting a C of A from the impurity vendor is usually pretty easy.

As a last resort, impurity standards can be characterized using chromatographic purity (area %) and moisture. It is a good idea to try to get some orthoganol methodology involved to be certain that you aren't missing any impurities in the impurity standards (MS, NMR, or at least a fairly different LC method w/ very different selectivity).

[Dan]

MarkFG,

As Sassman pointed out: RF = Area/Concentration

I have always expressed the concentration in units of 'mg/mL'. Other units may be used. The units used depend on the subsequent calculations used with the RF value.

There are two ways that I have used to determine the RF:
1) Make two preparations and six measurements each and calculate the mean RF.
2) Make a calibration curve. The Slope of the curve is the RF.

Do you intend to quantify your impurities using impurity standards or are you quantifying the impurities relative to the API standard response?

If you quantify the impurities relative to the API standard response, then you want the RRF (relative response factor) for each impurity. Calculate the RRF as:

RRF(imp) = Slope(API)/Slope(imp)

I would recommend using the RRF approach. Otherwise you will have to use impurity standards with each analysis and the impurity materials are usually in short supply and can be expensive. You only need to determine the RRF of each impurity once for the method (change the method and you may need to re-determine the RRF). Also, you would need (or may need) to qualify each batch/lot of impurity using the procedures described by DR. That can be expensive to do for each batch. So, the RRF approach is usually used.

Regards,
Dan

[gbalock]

We try to get the purest impurities as possible. For purity, the minimum information we use is chromatographic purity and moisture content. One impurity that we had used a sodium salt in its synthesis and we analyzed for Na. The purity of the impurity is taken into account if the purity is below 95%.

As far as procedure, we do a curve of the API and the impurity from slightly above LOQ to about 1% of the API in g/mL. We plot the response versus concetration and we divide the slope of the impurity to the slope of the API to get teh RRF.

[wanda50]

We use RRF since many of our impurities are hard to isolate and/or synthesize. Try to get LOD/LOQ for the impurity since customers often check for this when a method calls for RRF to be used.

In the calculation of RRF, you need to specify how the value is derived. We have a minimum of 5 data collection systems, and some use the RRF with the standard in the numerator, the others the analyte is in the numerator. So, make sure you know how your data collection system uses the value for calculations.

[Dan]

Wanda,

You bring up a good point about the RRF calculation. You can see from my earlier post and the post from George Blalock that we each are calculating the RFF differently.

There was a stimuli article in the USP PF about a year back that recommended standardization for the RRF, both in terminology and in calculation. It seems that a response factor of some type is used in many monographs but there is no consistency as to the term used (RF, CF, RRF, etc.) or the manner of the calculation. I don't have the article, I will need to look up the reference. I don't know the current status of this in the USP.

In the CDS software that I have used, the RRF term is a "multiplier" in the CDS program. So, I have used the RRF equation as I described in my previous post.

MarkFG,

One thing to add from my previous post. If the RRF is much different than 1.0, then an RRF should not be used. If the RRF is less than 0.05 or greater than 2.0 then you shouldn't use RRF, use a standard instead (I think that those are the values that we have used). Also, if the RRF value is >= 0.8 and <=1.2, then you can use 1.0 as the RRF value.

Regards,
Dan

[Oliver]

Hi Dan,
I came across that Pharmacopoeial Forum article and it quite shocked me so much that in 2 days I reviewed all our methods' validation reports and actually I found one whose terminology was inconsistent with the others. (luckily there were no impacts as its use in the calculation was clearly specified)

Incidentally we term RRF as Relative Response Factor in a way as the impurity RF relative to the active RF RRF=RF(imp)/RF(active)

Then we use the reciprocal of the RRF as a Correction Factor. This then we use a a multiplier with the area of the impurity in question such that we 'normalise' it to that equivalent to the active, in a way correct it before quantification.

Just a small matter of terminology I guess as long as it is clearly defined within an organisation then it should be fine.